modelId
stringlengths 5
139
| author
stringlengths 2
42
| last_modified
timestamp[us, tz=UTC]date 2020-02-15 11:33:14
2025-06-03 12:30:42
| downloads
int64 0
223M
| likes
int64 0
11.7k
| library_name
stringclasses 466
values | tags
sequencelengths 1
4.05k
| pipeline_tag
stringclasses 54
values | createdAt
timestamp[us, tz=UTC]date 2022-03-02 23:29:04
2025-06-03 12:30:29
| card
stringlengths 11
1.01M
|
|---|---|---|---|---|---|---|---|---|---|
huggingtweets/behemilf | huggingtweets | 2021-06-23T19:06:29Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-02T23:29:05Z | ---
language: en
thumbnail: https://github.com/borisdayma/huggingtweets/blob/master/img/logo.png?raw=true
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1404753773939990533/2Ol60_sO_400x400.jpg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">Mom</div>
<div style="text-align: center; font-size: 14px;">@behemilf</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from Mom.
| Data | Mom |
| --- | --- |
| Tweets downloaded | 3241 |
| Retweets | 858 |
| Short tweets | 346 |
| Tweets kept | 2037 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/34zvujdl/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @behemilf's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/ss8n55dy) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/ss8n55dy/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/behemilf')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
huggingtweets/caelan_hudson | huggingtweets | 2021-06-23T18:55:51Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-02T23:29:05Z | ---
language: en
thumbnail: https://github.com/borisdayma/huggingtweets/blob/master/img/logo.png?raw=true
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1400205166763122689/Zjyw9G_i_400x400.jpg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">Caelan Hudson</div>
<div style="text-align: center; font-size: 14px;">@caelan_hudson</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from Caelan Hudson.
| Data | Caelan Hudson |
| --- | --- |
| Tweets downloaded | 1768 |
| Retweets | 696 |
| Short tweets | 139 |
| Tweets kept | 933 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/vrzri0az/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @caelan_hudson's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/2u9374qr) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/2u9374qr/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/caelan_hudson')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
huggingtweets/alexisuwualexis | huggingtweets | 2021-06-23T18:49:20Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-02T23:29:05Z | ---
language: en
thumbnail: https://www.huggingtweets.com/alexisuwualexis/1624474156240/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1337389555863982083/GFu_etbo_400x400.jpg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">Alexis (she/her) 🏳️⚧️</div>
<div style="text-align: center; font-size: 14px;">@alexisuwualexis</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from Alexis (she/her) 🏳️⚧️.
| Data | Alexis (she/her) 🏳️⚧️ |
| --- | --- |
| Tweets downloaded | 3219 |
| Retweets | 2988 |
| Short tweets | 64 |
| Tweets kept | 167 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/t0aheh4s/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @alexisuwualexis's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/18q8udnh) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/18q8udnh/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/alexisuwualexis')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
huggingtweets/cookie__sophie | huggingtweets | 2021-06-23T18:38:16Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-02T23:29:05Z | ---
language: en
thumbnail: https://www.huggingtweets.com/cookie__sophie/1624473491534/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1385160467778310144/WyzPNrHb_400x400.jpg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">🐱Sophie/Cookie🍪🏳️⚧️</div>
<div style="text-align: center; font-size: 14px;">@cookie__sophie</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from 🐱Sophie/Cookie🍪🏳️⚧️.
| Data | 🐱Sophie/Cookie🍪🏳️⚧️ |
| --- | --- |
| Tweets downloaded | 3232 |
| Retweets | 463 |
| Short tweets | 375 |
| Tweets kept | 2394 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/15ifdxlx/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @cookie__sophie's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/390kytab) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/390kytab/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/cookie__sophie')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
huggingtweets/sbubby4 | huggingtweets | 2021-06-23T18:37:07Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-02T23:29:05Z | ---
language: en
thumbnail: https://www.huggingtweets.com/sbubby4/1624473423478/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1399079285411954690/Luvg7-oO_400x400.jpg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">sword witch</div>
<div style="text-align: center; font-size: 14px;">@sbubby4</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from sword witch.
| Data | sword witch |
| --- | --- |
| Tweets downloaded | 3214 |
| Retweets | 393 |
| Short tweets | 65 |
| Tweets kept | 2756 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/29ai7ons/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @sbubby4's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/k25px1ln) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/k25px1ln/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/sbubby4')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
huggingtweets/roedeerrootie | huggingtweets | 2021-06-23T18:36:45Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-02T23:29:05Z | ---
language: en
thumbnail: https://www.huggingtweets.com/roedeerrootie/1624473381138/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1399885746392092675/_GRuvCla_400x400.jpg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">Rootie</div>
<div style="text-align: center; font-size: 14px;">@roedeerrootie</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from Rootie.
| Data | Rootie |
| --- | --- |
| Tweets downloaded | 3209 |
| Retweets | 902 |
| Short tweets | 317 |
| Tweets kept | 1990 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/p726kemt/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @roedeerrootie's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/2my39bl0) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/2my39bl0/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/roedeerrootie')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
huggingtweets/reverse_city | huggingtweets | 2021-06-23T18:34:00Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-02T23:29:05Z | ---
language: en
thumbnail: https://www.huggingtweets.com/reverse_city/1624473236292/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1407628793188061186/du6ZO2Qz_400x400.jpg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">chloe</div>
<div style="text-align: center; font-size: 14px;">@reverse_city</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from chloe.
| Data | chloe |
| --- | --- |
| Tweets downloaded | 3249 |
| Retweets | 538 |
| Short tweets | 1503 |
| Tweets kept | 1208 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/2op7i1vy/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @reverse_city's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/15jzq6d0) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/15jzq6d0/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/reverse_city')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
huggingtweets/alisonselby_ | huggingtweets | 2021-06-23T18:32:39Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-02T23:29:05Z | ---
language: en
thumbnail: https://www.huggingtweets.com/alisonselby_/1624473155604/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1406680256258482178/79-ZrVAg_400x400.jpg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">Alison Selby</div>
<div style="text-align: center; font-size: 14px;">@alisonselby_</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from Alison Selby.
| Data | Alison Selby |
| --- | --- |
| Tweets downloaded | 3218 |
| Retweets | 319 |
| Short tweets | 290 |
| Tweets kept | 2609 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/2e6i4sab/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @alisonselby_'s tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/9gpt8ktz) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/9gpt8ktz/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/alisonselby_')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
huggingtweets/axel_hugsky | huggingtweets | 2021-06-23T18:30:11Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-02T23:29:05Z | ---
language: en
thumbnail: https://www.huggingtweets.com/axel_hugsky/1624473007421/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1402029332516773888/oJJ69stf_400x400.jpg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">Axel! ♠️</div>
<div style="text-align: center; font-size: 14px;">@axel_hugsky</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from Axel! ♠️.
| Data | Axel! ♠️ |
| --- | --- |
| Tweets downloaded | 3244 |
| Retweets | 529 |
| Short tweets | 1491 |
| Tweets kept | 1224 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/2ox7p0bd/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @axel_hugsky's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/rrwwxdal) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/rrwwxdal/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/axel_hugsky')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
huggingtweets/shishibane | huggingtweets | 2021-06-23T18:24:55Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-02T23:29:05Z | ---
language: en
thumbnail: https://www.huggingtweets.com/shishibane/1624472691094/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/1387047792321785868/uKccHxMl_400x400.jpg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">ShiShibane</div>
<div style="text-align: center; font-size: 14px;">@shishibane</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from ShiShibane.
| Data | ShiShibane |
| --- | --- |
| Tweets downloaded | 1053 |
| Retweets | 115 |
| Short tweets | 208 |
| Tweets kept | 730 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/1je8s399/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @shishibane's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/bye9hdkq) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/bye9hdkq/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/shishibane')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
Pollawat/mt5-small-thai-qg | Pollawat | 2021-06-23T14:57:30Z | 17 | 0 | transformers | [
"transformers",
"pytorch",
"mt5",
"text2text-generation",
"question-generation",
"dataset:NSC2018",
"license:mit",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z | ---
tags:
- question-generation
language:
- thai
- th
datasets:
- NSC2018
license: mit
---
[Google's mT5](https://github.com/google-research/multilingual-t5)
This is a model for generating questions from Thai texts. It was fine-tuned on NSC2018 corpus
```python
from transformers import T5Tokenizer, MT5ForConditionalGeneration
tokenizer = T5Tokenizer.from_pretrained("Pollawat/mt5-small-thai-qg")
model = MT5ForConditionalGeneration.from_pretrained("Pollawat/mt5-small-thai-qg")
text = "กรุงเทพมหานคร เป็นเมืองหลวงและนครที่มีประชากรมากที่สุดของประเทศไทย เป็นศูนย์กลางการปกครอง การศึกษา การคมนาคมขนส่ง การเงินการธนาคาร การพาณิชย์ การสื่อสาร และความเจริญของประเทศ เป็นเมืองที่มีชื่อยาวที่สุดในโลก ตั้งอยู่บนสามเหลี่ยมปากแม่น้ำเจ้าพระยา มีแม่น้ำเจ้าพระยาไหลผ่านและแบ่งเมืองออกเป็น 2 ฝั่ง คือ ฝั่งพระนครและฝั่งธนบุรี กรุงเทพมหานครมีพื้นที่ทั้งหมด 1,568.737 ตร.กม. มีประชากรตามทะเบียนราษฎรกว่า 5 ล้านคน ทำให้กรุงเทพมหานครเป็นเอกนคร (Primate City) จัด มีผู้กล่าวว่า กรุงเทพมหานครเป็น 'เอกนครที่สุดในโลก' เพราะมีประชากรมากกว่านครที่มีประชากรมากเป็นอันดับ 2 ถึง 40 เท่า[3]"
input_ids = tokenizer.encode(text, return_tensors='pt')
beam_output = model.generate(
input_ids,
max_length=50,
num_beams=5,
early_stopping=True
)
print(tokenizer.decode(beam_output[0], skip_special_tokens=True))
>> <extra_id_0>ของกรุงเทพมหานครเป็นเมืองหลวงของประเทศใด
``` |
Davlan/mt5_base_yor_eng_mt | Davlan | 2021-06-23T14:51:23Z | 9 | 0 | transformers | [
"transformers",
"pytorch",
"mt5",
"text2text-generation",
"arxiv:2103.08647",
"autotrain_compatible",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z | Hugging Face's logo
---
language:
- yo
- en
datasets:
- JW300 + [Menyo-20k](https://huggingface.co/datasets/menyo20k_mt)
---
# mT5_base_yor_eng_mt
## Model description
**mT5_base_yor_eng_mt** is a **machine translation** model from Yorùbá language to English language based on a fine-tuned mT5-base model. It establishes a **strong baseline** for automatically translating texts from Yorùbá to English.
Specifically, this model is a *mT5_base* model that was fine-tuned on JW300 Yorùbá corpus and [Menyo-20k](https://huggingface.co/datasets/menyo20k_mt)
## Intended uses & limitations
#### How to use
You can use this model with Transformers *pipeline* for MT.
```python
from transformers import MT5ForConditionalGeneration, T5Tokenizer
model = MT5ForConditionalGeneration.from_pretrained("Davlan/mt5_base_yor_eng_mt")
tokenizer = T5Tokenizer.from_pretrained("google/mt5-base")
input_string = "Akọni ajìjàgbara obìnrin tó sun àtìmalé torí owó orí"
inputs = tokenizer.encode(input_string, return_tensors="pt")
generated_tokens = model.generate(inputs)
results = tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
print(results)
```
#### Limitations and bias
This model is limited by its training dataset of entity-annotated news articles from a specific span of time. This may not generalize well for all use cases in different domains.
## Training data
This model was fine-tuned on on JW300 Yorùbá corpus and [Menyo-20k](https://huggingface.co/datasets/menyo20k_mt) dataset
## Training procedure
This model was trained on a single NVIDIA V100 GPU
## Eval results on Test set (BLEU score)
15.57 BLEU on [Menyo-20k test set](https://arxiv.org/abs/2103.08647)
### BibTeX entry and citation info
By David Adelani
```
```
|
valhalla/t5-small-qa-qg-hl | valhalla | 2021-06-23T14:42:41Z | 2,105 | 12 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"question-generation",
"dataset:squad",
"arxiv:1910.10683",
"license:mit",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:05Z | ---
datasets:
- squad
tags:
- question-generation
widget:
- text: "generate question: <hl> 42 <hl> is the answer to life, the universe and everything. </s>"
- text: "question: What is 42 context: 42 is the answer to life, the universe and everything. </s>"
license: mit
---
## T5 for multi-task QA and QG
This is multi-task [t5-small](https://arxiv.org/abs/1910.10683) model trained for question answering and answer aware question generation tasks.
For question generation the answer spans are highlighted within the text with special highlight tokens (`<hl>`) and prefixed with 'generate question: '. For QA the input is processed like this `question: question_text context: context_text </s>`
You can play with the model using the inference API. Here's how you can use it
For QG
`generate question: <hl> 42 <hl> is the answer to life, the universe and everything. </s>`
For QA
`question: What is 42 context: 42 is the answer to life, the universe and everything. </s>`
For more deatils see [this](https://github.com/patil-suraj/question_generation) repo.
### Model in action 🚀
You'll need to clone the [repo](https://github.com/patil-suraj/question_generation).
[](https://colab.research.google.com/github/patil-suraj/question_generation/blob/master/question_generation.ipynb)
```python3
from pipelines import pipeline
nlp = pipeline("multitask-qa-qg")
# to generate questions simply pass the text
nlp("42 is the answer to life, the universe and everything.")
=> [{'answer': '42', 'question': 'What is the answer to life, the universe and everything?'}]
# for qa pass a dict with "question" and "context"
nlp({
"question": "What is 42 ?",
"context": "42 is the answer to life, the universe and everything."
})
=> 'the answer to life, the universe and everything'
``` |
valhalla/t5-base-qg-hl | valhalla | 2021-06-23T14:40:47Z | 5,139 | 11 | transformers | [
"transformers",
"pytorch",
"t5",
"text2text-generation",
"question-generation",
"dataset:squad",
"arxiv:1910.10683",
"license:mit",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:05Z | ---
datasets:
- squad
tags:
- question-generation
widget:
- text: "<hl> 42 <hl> is the answer to life, the universe and everything. </s>"
- text: "Python is a programming language. It is developed by <hl> Guido Van Rossum <hl>. </s>"
- text: "Although <hl> practicality <hl> beats purity </s>"
license: mit
---
## T5 for question-generation
This is [t5-base](https://arxiv.org/abs/1910.10683) model trained for answer aware question generation task. The answer spans are highlighted within the text with special highlight tokens.
You can play with the model using the inference API, just highlight the answer spans with `<hl>` tokens and end the text with `</s>`. For example
`<hl> 42 <hl> is the answer to life, the universe and everything. </s>`
For more deatils see [this](https://github.com/patil-suraj/question_generation) repo.
### Model in action 🚀
You'll need to clone the [repo](https://github.com/patil-suraj/question_generation).
[](https://colab.research.google.com/github/patil-suraj/question_generation/blob/master/question_generation.ipynb)
```python3
from pipelines import pipeline
nlp = pipeline("question-generation", model="valhalla/t5-base-qg-hl")
nlp("42 is the answer to life, universe and everything.")
=> [{'answer': '42', 'question': 'What is the answer to life, universe and everything?'}]
``` |
syndi-models/titlewave-t5-base | syndi-models | 2021-06-23T14:26:41Z | 8 | 1 | transformers | [
"transformers",
"pytorch",
"t5",
"text2text-generation",
"summarization",
"en",
"license:cc-by-4.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2023-05-09T20:00:31Z | ---
language: en
license: cc-by-4.0
pipeline_tag: summarization
widget:
- text: "Example question body."
---
# Titlewave: t5-base
## Model description
Titlewave is a Chrome extension that helps you choose better titles for your Stack Overflow questions. See https://github.com/tennessejoyce/TitleWave for more information.
This is one of two NLP models used in the Titlewave project, and its purpose is to suggests a new title based on on the body of the question. The companion model (https://huggingface.co/tennessejoyce/titlewave-bert-base-uncased) classifies whether question will be answered or not just based on the title
## Intended use
Try out different titles for your Stack Overflow post, and see which one gives you the best chance of recieving an answer.
This model can be used in your browser as a Chrome extension by following the installation instructions at https://github.com/tennessejoyce/TitleWave.
Or load it in Python like this (which will automatically download the model to your machine):
```python
>>> from transformers import pipeline
>>> classifier = pipeline('summarization', model='tennessejoyce/titlewave-t5-base')
>>> body = """"Example question body."""
>>> classifier(body)
[{'summary_text': 'Example title suggestion?'}]
```
## Training data
The weights were initialized from the BERT base model (https://huggingface.co/bert-base-uncased), which was trained on BookCorpus and English Wikipedia.
Then the model was fine-tuned on the dataset of previous Stack Overflow post titles (https://archive.org/details/stackexchange).
Specifically I used three years of posts from 2017-2019, filtered out posts which were closed, and selected 25% of the remaining posts at random to use in
the training set. In order to improve the quality of the titles generated, the model was trained only on questions with an accepted answer.
## Evaluation
See https://github.com/tennessejoyce/TitleWave/blob/master/model_training/test_summarizer.ipynb for the performance of the title generation model on the test set.
|
tennessejoyce/titlewave-t5-base | tennessejoyce | 2021-06-23T14:26:41Z | 20 | 5 | transformers | [
"transformers",
"pytorch",
"t5",
"text2text-generation",
"summarization",
"en",
"license:cc-by-4.0",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:05Z | ---
language: en
license: cc-by-4.0
pipeline_tag: summarization
widget:
- text: "Example question body."
---
# Titlewave: t5-base
## Model description
Titlewave is a Chrome extension that helps you choose better titles for your Stack Overflow questions. See https://github.com/tennessejoyce/TitleWave for more information.
This is one of two NLP models used in the Titlewave project, and its purpose is to suggests a new title based on on the body of the question. The companion model (https://huggingface.co/tennessejoyce/titlewave-bert-base-uncased) classifies whether question will be answered or not just based on the title
## Intended use
Try out different titles for your Stack Overflow post, and see which one gives you the best chance of recieving an answer.
This model can be used in your browser as a Chrome extension by following the installation instructions at https://github.com/tennessejoyce/TitleWave.
Or load it in Python like this (which will automatically download the model to your machine):
```python
>>> from transformers import pipeline
>>> classifier = pipeline('summarization', model='tennessejoyce/titlewave-t5-base')
>>> body = """"Example question body."""
>>> classifier(body)
[{'summary_text': 'Example title suggestion?'}]
```
## Training data
The weights were initialized from the BERT base model (https://huggingface.co/bert-base-uncased), which was trained on BookCorpus and English Wikipedia.
Then the model was fine-tuned on the dataset of previous Stack Overflow post titles (https://archive.org/details/stackexchange).
Specifically I used three years of posts from 2017-2019, filtered out posts which were closed, and selected 25% of the remaining posts at random to use in
the training set. In order to improve the quality of the titles generated, the model was trained only on questions with an accepted answer.
## Evaluation
See https://github.com/tennessejoyce/TitleWave/blob/master/model_training/test_summarizer.ipynb for the performance of the title generation model on the test set.
|
snrspeaks/t5-one-line-summary | snrspeaks | 2021-06-23T14:20:22Z | 1,385 | 91 | transformers | [
"transformers",
"pytorch",
"t5",
"text2text-generation",
"dataset:arxiv",
"license:mit",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:05Z | ---
datasets:
- arxiv
widget:
- text: "summarize: We describe a system called Overton, whose main design goal is to support engineers in building, monitoring, and improving production
machinelearning systems. Key challenges engineers face are monitoring fine-grained quality, diagnosing errors in sophisticated applications, and
handling contradictory or incomplete supervision data. Overton automates the life cycle of model construction, deployment, and monitoring by providing a set of novel high-level, declarative abstractions. Overton's vision is to shift developers to these higher-level tasks instead of lower-level machine learning tasks.
In fact, using Overton, engineers can build deep-learning-based applications without writing any code in frameworks like TensorFlow. For over a year,
Overton has been used in production to support multiple applications in both near-real-time applications and back-of-house processing.
In that time, Overton-based applications have answered billions of queries in multiple languages and processed trillions of records reducing errors
1.7-2.9 times versus production systems."
license: mit
---
# T5 One Line Summary
A T5 model trained on 370,000 research papers, to generate one line summary based on description/abstract of the papers. It is trained using [simpleT5](https://github.com/Shivanandroy/simpleT5) library - A python package built on top of pytorch lightning⚡️ & transformers🤗 to quickly train T5 models
## Usage:[](https://colab.research.google.com/drive/1HrfT8IKLXvZzPFpl1EhZ3s_iiXG3O2VY?usp=sharing)
```python
abstract = """We describe a system called Overton, whose main design goal is to support engineers in building, monitoring, and improving production
machine learning systems. Key challenges engineers face are monitoring fine-grained quality, diagnosing errors in sophisticated applications, and
handling contradictory or incomplete supervision data. Overton automates the life cycle of model construction, deployment, and monitoring by providing a
set of novel high-level, declarative abstractions. Overton's vision is to shift developers to these higher-level tasks instead of lower-level machine learning tasks.
In fact, using Overton, engineers can build deep-learning-based applications without writing any code in frameworks like TensorFlow. For over a year,
Overton has been used in production to support multiple applications in both near-real-time applications and back-of-house processing. In that time,
Overton-based applications have answered billions of queries in multiple languages and processed trillions of records reducing errors 1.7-2.9 times versus production systems.
"""
```
### Using Transformers🤗
```python
model_name = "snrspeaks/t5-one-line-summary"
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name)
input_ids = tokenizer.encode("summarize: " + abstract, return_tensors="pt", add_special_tokens=True)
generated_ids = model.generate(input_ids=input_ids,num_beams=5,max_length=50,repetition_penalty=2.5,length_penalty=1,early_stopping=True,num_return_sequences=3)
preds = [tokenizer.decode(g, skip_special_tokens=True, clean_up_tokenization_spaces=True) for g in generated_ids]
print(preds)
# output
["Overton: Building, Deploying, and Monitoring Machine Learning Systems for Engineers",
"Overton: A System for Building, Monitoring, and Improving Production Machine Learning Systems",
"Overton: Building, Monitoring, and Improving Production Machine Learning Systems"]
```
### Using simpleT5⚡️
```python
# pip install --upgrade simplet5
from simplet5 import SimpleT5
model = SimpleT5()
model.load_model("t5","snrspeaks/t5-one-line-summary")
model.predict(abstract)
# output
"Overton: Building, Deploying, and Monitoring Machine Learning Systems for Engineers"
``` |
prithivida/active_to_passive_styletransfer | prithivida | 2021-06-23T13:43:58Z | 123 | 4 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:05Z | ## This model belongs to the Styleformer project
[Please refer to github page](https://github.com/PrithivirajDamodaran/Styleformer)
|
manueldeprada/t5-cord19-paraphrase-paws-msrp-opinosis | manueldeprada | 2021-06-23T12:34:22Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:05Z | # T5-Paraphrase pretrained using the CORD-19 dataset.
The base model is manueldeprada/t5-cord19, which has been pretrained with the text and abstracts from the CORD-19 dataset.
It has been finetuned in paraphrasing text like ceshine/t5-paraphrase-paws-msrp-opinosis, using the scripts from [ceshine/finetuning-t5 Github repo](https://github.com/ceshine/finetuning-t5/tree/master/paraphrase).
It does the same paraphrasing but the CORD-19 pretraining allows this model to perform well in COVID-19 related text. |
dbernsohn/algebra_linear_1d_composed | dbernsohn | 2021-06-23T12:16:42Z | 7 | 0 | transformers | [
"transformers",
"pytorch",
"t5",
"text2text-generation",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:05Z | # algebra_linear_1d_composed
---
language: en
datasets:
- algebra_linear_1d_composed
---
This is a [t5-small](https://ai.googleblog.com/2020/02/exploring-transfer-learning-with-t5.html) fine-tuned version on the [math_dataset/algebra_linear_1d_composed](https://www.tensorflow.org/datasets/catalog/math_dataset#mathdatasetalgebra_linear_1d_composed) for solving **algebra linear 1d composed equations** mission.
To load the model:
(necessary packages: !pip install transformers sentencepiece)
```python
from transformers import AutoTokenizer, AutoModelWithLMHead
tokenizer = AutoTokenizer.from_pretrained("dbernsohn/algebra_linear_1d_composed")
model = AutoModelWithLMHead.from_pretrained("dbernsohn/algebra_linear_1d_composed")
```
You can then use this model to solve algebra 1d equations into numbers.
```python
query = "Suppose -d = 5 - 16. Let b = -579 + 584. Solve -b*c + 36 = d for c."
input_text = f"{query} </s>"
features = tokenizer([input_text], return_tensors='pt')
model.to('cuda')
output = model.generate(input_ids=features['input_ids'].cuda(),
attention_mask=features['attention_mask'].cuda())
tokenizer.decode(output[0])
# <pad> 5</s>
```
Another examples:
+ Suppose -d = 5 - 16. Let b = -579 + 584. Solve -b*c + 36 = d for c.
+ Answer: 5 Pred: 5
----
+ Suppose 3*v - l + 9 = 4*v, 0 = -5*v + 5*l - 5. Let f(s) = 3*s**2 + 1. Let g be f(-1). Suppose 63 = g*x - x. Solve -5*i + v + x = 0 for i.
+ Answer: 5 Pred: 5
----
+ Let w be 2 - (0 - 0)/(-2). Let f = -110 - -110. Suppose f*m - 4*m + 3*m = 0. Solve m*v = -w*v for v.
+ Answer: 0 Pred: 0
----
+ Let a(h) = -34*h**3 - 15 + 3*h + 36*h**3 + 8*h**2 + 5*h**2. Let r be a(-6). Solve 2*z = r*z for z.
+ Answer: 0 Pred: 0
----
+ Suppose -3*p + 24 = -3*c, 0*c + 6 = -2*c. Suppose -67 = 4*i + 289. Let t = i + 94. Solve t = 2*y - p for y.
+ Answer: 5 Pred: 5
----
+ Let b = -36 + 53. Suppose -7*u - b = -73. Solve j + 3*j = -u for j.
+ Answer: -2 Pred: -2
----
+ Let h be 8*((-2)/2 + 14)*1. Let y = -101 + h. Solve y*p = -p for p.
+ Answer: 0 Pred: 0
----
+ Let b = 178 - 79. Let s be 9/(-1 - 2 - b/(-22)). Solve s = -k - k for k.
+ Answer: -3 Pred: -3
----
+ Suppose 31 = -4*z + 11, -3*k - 5*z - 22 = 0. Solve 23 = -11*p + k for p.
+ Answer: -2 Pred: -2
The whole training process and hyperparameters are in my [GitHub repo](https://github.com/DorBernsohn/CodeLM/tree/main/MathLM)
> Created by [Dor Bernsohn](https://www.linkedin.com/in/dor-bernsohn-70b2b1146/)
|
m3hrdadfi/hubert-base-persian-speech-gender-recognition | m3hrdadfi | 2021-06-23T12:16:09Z | 2,573 | 7 | transformers | [
"transformers",
"pytorch",
"hubert",
"audio",
"speech",
"speech-gender-recognition",
"fa",
"dataset:shemo",
"license:apache-2.0",
"endpoints_compatible",
"region:us"
] | null | 2022-03-02T23:29:05Z | ---
language: fa
datasets:
- shemo
tags:
- audio
- speech
- speech-gender-recognition
license: apache-2.0
---
# Emotion Recognition in Persian (fa) Speech using HuBERT
## How to use
### Requirements
```bash
# requirement packages
!pip install git+https://github.com/huggingface/datasets.git
!pip install git+https://github.com/huggingface/transformers.git
!pip install torchaudio
!pip install librosa
```
```bash
!git clone https://github.com/m3hrdadfi/soxan.git .
```
### Prediction
```python
import torch
import torch.nn as nn
import torch.nn.functional as F
import torchaudio
from transformers import AutoConfig, Wav2Vec2FeatureExtractor
from src.models import Wav2Vec2ForSpeechClassification, HubertForSpeechClassification
import librosa
import IPython.display as ipd
import numpy as np
import pandas as pd
```
```python
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model_name_or_path = "m3hrdadfi/hubert-base-persian-speech-gender-recognition"
config = AutoConfig.from_pretrained(model_name_or_path)
feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained(model_name_or_path)
sampling_rate = feature_extractor.sampling_rate
model = HubertForSpeechClassification.from_pretrained(model_name_or_path).to(device)
```
```python
def speech_file_to_array_fn(path, sampling_rate):
speech_array, _sampling_rate = torchaudio.load(path)
resampler = torchaudio.transforms.Resample(_sampling_rate)
speech = resampler(speech_array).squeeze().numpy()
return speech
def predict(path, sampling_rate):
speech = speech_file_to_array_fn(path, sampling_rate)
inputs = feature_extractor(speech, sampling_rate=sampling_rate, return_tensors="pt", padding=True)
inputs = {key: inputs[key].to(device) for key in inputs}
with torch.no_grad():
logits = model(**inputs).logits
scores = F.softmax(logits, dim=1).detach().cpu().numpy()[0]
outputs = [{"Label": config.id2label[i], "Score": f"{round(score * 100, 3):.1f}%"} for i, score in enumerate(scores)]
return outputs
```
```python
path = "/path/to/female.wav"
outputs = predict(path, sampling_rate)
```
```bash
[{'Label': 'F', 'Score': '98.2%'}, {'Label': 'M', 'Score': '1.8%'}]
```
## Evaluation
The following tables summarize the scores obtained by model overall and per each class.
| Emotions | precision | recall | f1-score | accuracy |
|----------|-----------|--------|----------|----------|
| F | 0.98 | 0.97 | 0.98 | |
| M | 0.98 | 0.99 | 0.98 | |
| | | | Overal | 0.98 |
## Questions?
Post a Github issue from [HERE](https://github.com/m3hrdadfi/soxan/issues). |
castorini/t5-base-canard | castorini | 2021-06-23T11:56:05Z | 155 | 2 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:05Z | This model is trained for conversational question rewriting.
Usage:
Source text format: ${HISTORY} ||| ${CURRENT_QUESTION}
example from [CANARD](https://sites.google.com/view/qanta/projects/canard):
Frank Zappa ||| Disbandment ||| What group disbanded ||| Zappa and the Mothers of Invention ||| When did they disband?
Target text:
When did Zappa and the Mothers of Invention disband?
You can find our guide to reproduce the training in this [repo](https://github.com/castorini/chatty-goose/blob/c7d0cd8c45354b09b5fb930ab0b5af8be2e5772b/docs/t5_finetuning.md). |
castorini/monot5-base-med-msmarco | castorini | 2021-06-23T11:40:06Z | 9 | 1 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | feature-extraction | 2022-03-02T23:29:05Z | This model is a T5-base reranker fine-tuned on the MS MARCO passage dataset for 10k steps (or 1 epoch) and then fine-tuned again on MedMARCO (from [Sledge-Z paper](https://www.aclweb.org/anthology/2020.emnlp-main.341.pdf) for 1k steps.
For more details on how to use it, check [pygaggle.ai](pygaggle.ai)
Paper describing the model: [Document Ranking with a Pretrained Sequence-to-Sequence Model](https://www.aclweb.org/anthology/2020.findings-emnlp.63/) |
SEBIS/legal_t5_small_trans_fr_en_small_finetuned | SEBIS | 2021-06-23T11:38:04Z | 15 | 0 | transformers | [
"transformers",
"pytorch",
"t5",
"text2text-generation",
"translation French English model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: French English
tags:
- translation French English model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "RÉSULTAT DU VOTE FINAL EN COMMISSION"
---
# legal_t5_small_trans_fr_en_small_finetuned model
Model on translating legal text from French to English. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is first pretrained all the translation data over some unsupervised task. Then the model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_fr_en_small_finetuned is initially pretrained on unsupervised task with the all of the data of the training set. The unsupervised task was "masked language modelling". legal_t5_small_trans_fr_en_small_finetuned is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from French to English.
### How to use
Here is how to use this model to translate legal text from French to English in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_fr_en_small_finetuned"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_fr_en", do_lower_case=False,
skip_special_tokens=True),
device=0
)
fr_text = "RÉSULTAT DU VOTE FINAL EN COMMISSION"
pipeline([fr_text], max_length=512)
```
## Training data
The legal_t5_small_trans_fr_en_small_finetuned (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 9 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
The pre-training data was the combined data from all the 42 language pairs. The task for the model was to predict the portions of a sentence which were masked randomly.
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_fr_en_small_finetuned | 51.351|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_de_cs | SEBIS | 2021-06-23T11:37:25Z | 19 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Deustch Cszech model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Deustch Cszech
tags:
- translation Deustch Cszech model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "17. empfiehlt die Einführung einer spezifischen Strategie zur Unterstützung neuer und demokratisch gewählter Parlamente im Hinblick auf eine dauerhafte Verankerung von Demokratie, Rechtsstaatlichkeit und guter Staatsführung;"
---
# legal_t5_small_trans_de_cs model
Model on translating legal text from Deustch to Cszech. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_de_cs is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Deustch to Cszech.
### How to use
Here is how to use this model to translate legal text from Deustch to Cszech in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_de_cs"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_de_cs", do_lower_case=False,
skip_special_tokens=True),
device=0
)
de_text = "17. empfiehlt die Einführung einer spezifischen Strategie zur Unterstützung neuer und demokratisch gewählter Parlamente im Hinblick auf eine dauerhafte Verankerung von Demokratie, Rechtsstaatlichkeit und guter Staatsführung;"
pipeline([de_text], max_length=512)
```
## Training data
The legal_t5_small_trans_de_cs model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_de_cs | 44.07|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_cs_sv_small_finetuned | SEBIS | 2021-06-23T11:36:51Z | 9 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Cszech Swedish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech Swedish
tags:
- translation Cszech Swedish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "10 Ukončení denního zasedání"
---
# legal_t5_small_trans_cs_sv_small_finetuned model
Model on translating legal text from Cszech to Swedish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is first pretrained all the translation data over some unsupervised task. Then the model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_cs_sv_small_finetuned is initially pretrained on unsupervised task with the all of the data of the training set. The unsupervised task was "masked language modelling". legal_t5_small_trans_cs_sv_small_finetuned is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Cszech to Swedish.
### How to use
Here is how to use this model to translate legal text from Cszech to Swedish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_cs_sv_small_finetuned"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_cs_sv", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "10 Ukončení denního zasedání"
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_trans_cs_sv_small_finetuned (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
The pre-training data was the combined data from all the 42 language pairs. The task for the model was to predict the portions of a sentence which were masked randomly.
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_cs_sv_small_finetuned | 48.159|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_cs_it | SEBIS | 2021-06-23T11:35:03Z | 20 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Cszech Italian model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech Italian
tags:
- translation Cszech Italian model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "– Měly by se podporovat normy sportovní správy prostřednictvím výměny osvědčených postupů."
---
# legal_t5_small_trans_cs_it model
Model on translating legal text from Cszech to Italian. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_cs_it is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Cszech to Italian.
### How to use
Here is how to use this model to translate legal text from Cszech to Italian in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_cs_it"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_cs_it", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "– Měly by se podporovat normy sportovní správy prostřednictvím výměny osvědčených postupů."
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_trans_cs_it model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_cs_it | 46.67|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_cs_fr | SEBIS | 2021-06-23T11:33:48Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Cszech French model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech French
tags:
- translation Cszech French model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Prevencí proti nemoci Usnesení, o kterém bude Parlament hlasovat 24. října je založeno zejména na interpelacích, které poslancům předložily parlamentní kluby pro životní prostředí, zaměstnanost a práva žen."
---
# legal_t5_small_trans_cs_fr model
Model on translating legal text from Cszech to French. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_cs_fr is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Cszech to French.
### How to use
Here is how to use this model to translate legal text from Cszech to French in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_cs_fr"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_cs_fr", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "Prevencí proti nemoci Usnesení, o kterém bude Parlament hlasovat 24. října je založeno zejména na interpelacích, které poslancům předložily parlamentní kluby pro životní prostředí, zaměstnanost a práva žen."
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_trans_cs_fr model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_cs_fr | 50.75|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_cs_es_small_finetuned | SEBIS | 2021-06-23T11:32:56Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"t5",
"text2text-generation",
"translation Cszech Spanish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech Spanish
tags:
- translation Cszech Spanish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "vzhledem k tomu, že parlamentní volby v listopadu a v prosinci 2006, volby do Senátu v lednu 2007 a volbu prezidenta Sídí Muhammada Ulda Šajcha Abdalláhiho v březnu 2007, uznali jako spravedlivé a transparentní zahraniční pozorovatelé, včetně pozorovatelů z Evropské unie, a zejména z mise ke sledování průběhu voleb vyslané Evropským parlamentem, jenž se tím stal garantem legality těchto voleb,"
---
# legal_t5_small_trans_cs_es_small_finetuned model
Model on translating legal text from Cszech to Spanish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is first pretrained all the translation data over some unsupervised task. Then the model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_cs_es_small_finetuned is initially pretrained on unsupervised task with the all of the data of the training set. The unsupervised task was "masked language modelling". legal_t5_small_trans_cs_es_small_finetuned is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Cszech to Spanish.
### How to use
Here is how to use this model to translate legal text from Cszech to Spanish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_cs_es_small_finetuned"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_cs_es", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "vzhledem k tomu, že parlamentní volby v listopadu a v prosinci 2006, volby do Senátu v lednu 2007 a volbu prezidenta Sídí Muhammada Ulda Šajcha Abdalláhiho v březnu 2007, uznali jako spravedlivé a transparentní zahraniční pozorovatelé, včetně pozorovatelů z Evropské unie, a zejména z mise ke sledování průběhu voleb vyslané Evropským parlamentem, jenž se tím stal garantem legality těchto voleb,"
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_trans_cs_es_small_finetuned (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
The pre-training data was the combined data from all the 42 language pairs. The task for the model was to predict the portions of a sentence which were masked randomly.
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_cs_es_small_finetuned | 50.862|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_cs_es | SEBIS | 2021-06-23T11:32:25Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Cszech Spanish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech Spanish
tags:
- translation Cszech Spanish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "k návrhu směrnice Evropského parlamentu a Rady o bezpečnosti hraček"
---
# legal_t5_small_trans_cs_es model
Model on translating legal text from Cszech to Spanish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_cs_es is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Cszech to Spanish.
### How to use
Here is how to use this model to translate legal text from Cszech to Spanish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_cs_es"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_cs_es", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "k návrhu směrnice Evropského parlamentu a Rady o bezpečnosti hraček"
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_trans_cs_es model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_cs_es | 50.77|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_cs_en_small_finetuned | SEBIS | 2021-06-23T11:31:44Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Cszech English model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech English
tags:
- translation Cszech English model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "4) Seznam užívaných výrobků s obsahem PFOS: Kvůli značnému poklesu výroby PFOS po roce 2000 představují největší zdroj emisí patrně dřívější využití, která však nadále reálně existují."
---
# legal_t5_small_trans_cs_en_small_finetuned model
Model on translating legal text from Cszech to English. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is first pretrained all the translation data over some unsupervised task. Then the model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_cs_en_small_finetuned is initially pretrained on unsupervised task with the all of the data of the training set. The unsupervised task was "masked language modelling". legal_t5_small_trans_cs_en_small_finetuned is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Cszech to English.
### How to use
Here is how to use this model to translate legal text from Cszech to English in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_cs_en_small_finetuned"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_cs_en", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "4) Seznam užívaných výrobků s obsahem PFOS: Kvůli značnému poklesu výroby PFOS po roce 2000 představují největší zdroj emisí patrně dřívější využití, která však nadále reálně existují."
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_trans_cs_en_small_finetuned (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
The pre-training data was the combined data from all the 42 language pairs. The task for the model was to predict the portions of a sentence which were masked randomly.
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_cs_en_small_finetuned | 56.936|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_cs_en | SEBIS | 2021-06-23T11:30:54Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Cszech English model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech English
tags:
- translation Cszech English model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "s ohledem na druhou schůzku států OSN, která se konala 11.–15. června 2005 a měla posoudit provádění akčního programu OSN k prevenci, potírání a vymýcení nezákonného obchodu s ručními a lehkými zbraněmi ve všech jeho aspektech, která se koná jednou za dva roky,"
---
# legal_t5_small_trans_cs_en model
Model on translating legal text from Cszech to English. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_cs_en is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Cszech to English.
### How to use
Here is how to use this model to translate legal text from Cszech to English in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_cs_en"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_cs_en", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "s ohledem na druhou schůzku států OSN, která se konala 11.–15. června 2005 a měla posoudit provádění akčního programu OSN k prevenci, potírání a vymýcení nezákonného obchodu s ručními a lehkými zbraněmi ve všech jeho aspektech, která se koná jednou za dva roky,"
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_trans_cs_en model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_cs_en | 56.92|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_cs_de_small_finetuned | SEBIS | 2021-06-23T11:30:18Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Cszech Deustch model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech Deustch
tags:
- translation Cszech Deustch model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Vzhledem k tomu, že tento právní předpis bude přímo použitelný v členských státech a zavede mnoho povinností pro ty, na něž se vztahuje, je žádoucí, aby se jim poskytlo více času na přizpůsobení se těmto novým pravidlům."
---
# legal_t5_small_trans_cs_de_small_finetuned model
Model on translating legal text from Cszech to Deustch. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is first pretrained all the translation data over some unsupervised task. Then the model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_cs_de_small_finetuned is initially pretrained on unsupervised task with the all of the data of the training set. The unsupervised task was "masked language modelling". legal_t5_small_trans_cs_de_small_finetuned is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Cszech to Deustch.
### How to use
Here is how to use this model to translate legal text from Cszech to Deustch in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_cs_de_small_finetuned"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_cs_de", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "Vzhledem k tomu, že tento právní předpis bude přímo použitelný v členských státech a zavede mnoho povinností pro ty, na něž se vztahuje, je žádoucí, aby se jim poskytlo více času na přizpůsobení se těmto novým pravidlům."
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_trans_cs_de_small_finetuned (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
The pre-training data was the combined data from all the 42 language pairs. The task for the model was to predict the portions of a sentence which were masked randomly.
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_cs_de_small_finetuned | 44.175|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_cs_de | SEBIS | 2021-06-23T11:29:34Z | 9 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Cszech Deustch model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech Deustch
tags:
- translation Cszech Deustch model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Konečná zpráva bude Parlamentu předložena na konci nového funkčního období."
---
# legal_t5_small_trans_cs_de model
Model on translating legal text from Cszech to Deustch. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_cs_de is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Cszech to Deustch.
### How to use
Here is how to use this model to translate legal text from Cszech to Deustch in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_cs_de"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_cs_de", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "Konečná zpráva bude Parlamentu předložena na konci nového funkčního období."
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_trans_cs_de model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_cs_de | 44.69|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
auday/paraphraser_model1 | auday | 2021-06-23T11:29:03Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:05Z | This folder contain a Google T5 Transformer Fine-tuned to generate paraphrases using:
- Para_NMT_50M_Paraphrasing_train_small.csv 134337 lines of pair sentences 19Mbytes
- Para_NMT_50M_Paraphrasing_val_small.csv 14928 lines of pair sentences 2.0Mbytes
Training Start Time: Sun Mar 14 18:27:15 2021
Training End Time: Sun Mar 14 22:19:00 2021
|
SEBIS/legal_t5_small_summ_sv | SEBIS | 2021-06-23T11:28:45Z | 6 | 1 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"summarization Swedish model",
"dataset:jrc-acquis",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Swedish
tags:
- summarization Swedish model
datasets:
- jrc-acquis
widget:
- text: "EUROPEISKA GEMENSKAPERNAS RÅD HAR ANTAGIT DENNA FÖRORDNING med beaktande av Fördraget om upprättandet av Europeiska ekonomiska gemenskapen, särskilt artiklarna 43 och 100a i detta, med beaktande av kommissionens förslag(1), i samarbete med Europaparlamentet(2), med beaktande av Ekonomiska och sociala kommitténs yttrande(3), och med beaktande av följande: Det bör införas förbud mot användning av blybaserade kapsyler eller blybaserad folie i förslutningar på förpackningar som används då aromatiserade viner, aromatiserade vinbaserade drycker och aromatiserade drinkar baserade på vinprodukter släpps ut på marknaden i syfte att undvika risken för kontaminering, särskilt vid oavsiktlig kontakt med sådana produkter, samt risken för miljöförorening på grund av avfall som innehåller bly från kapsyler och folie av detta slag. Tillverkarna och användarna av kapsylerna och folien i fråga bör dock ges tid att anpassa sig genom att förbudet inte tillämpas förrän från och med den 1 januari 1993. Det är även nödvändigt att tillåta att produkter som före detta datum tappats på buteljer med blybaserade kapsyler eller blybaserad folie får säljas till dess att lagren är uttömda. Vissa definitioner av aromatiserade vinbaserade drycker bör anpassas så att större hänsyn tas till traditionella framställningsmetoder. Förordning (EEG) nr 1601/91(4) bör därför ändras. HÄRIGENOM FÖRESKRIVS FÖLJANDE. Artikel 1 Förordning (EEG) nr 1601/91 ändras på följande sätt: 1. Artikel 2.3 a första stycket skall ersättas med följande: %quot%a) Sangria: en dryck som framställs av vin - som smaksatts genom tillsats av naturliga extrakt eller essenser av citrusfrukt, - med eller utan saft av sådan frukt, - eventuellt: - med tillsats av kryddor, - sötat, - med tillsats av CO2, och med en slutlig alkoholstyrka på under 12 volymprocent.%quot% 2. Artikel 2.3 e skall ersättas med följande: %quot%e) Kalte Ente: Smaksatt vinbaserad dryck som framställs genom att vin, pärlande vin eller pärlande vin med tillsatt CO2 blandas med mousserande vin eller mousserande vin med tillsatt CO2 och tillsätts naturlig citronsubstans eller extrakt av detta som måste ge en tydligt framträdande smak. Slutprodukten måste innehålla minst 25 volymprocent mousserande vin eller mousserande vin med tillsatt CO2.%quot% 3. Följande punkt skall införas i artikel 8: %quot%4.a Från och med den 1 januari 1993 får buteljerade produkter som omfattas av denna förordning inte saluhållas eller släppas ut på marknaden i förpackningar med förslutningar som täckts med blybaserade kapsyler eller blybaserad folie. Dock får produkter som före detta datum tappats på flaskor med detta slag av kapsyler eller folie avyttras till dess att lagren tömts.%quot% Artikel 2 Denna förordning träder i kraft den tredje dagen efter det att den har offentliggjorts i Europeiska gemenskapernas officiella tidning. Denna förordning är till alla delar bindande och direkt tillämplig i alla medlemsstater. Utfärdad i Bryssel den 9 november 1992. På rådets vägnar D. HURD Ordförande (1) EGT nr C 69, 18.3.1992, s. 11. (2) EGT nr C 241, 21.9.1992, s. 97 och beslut av den 28 oktober 1992. (3) EGT nr C 169, 6.7.1992, s. 1. (4) EGT nr L 149, 14.6.1991, s. 1. "
---
# legal_t5_small_summ_sv model
Model for Summarization of legal text written in Swedish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis.
## Model description
legal_t5_small_summ_sv is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for summarization of legal texts written in Swedish.
### How to use
Here is how to use this model to summarize legal text written in Swedish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_summ_sv"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_summ_sv", do_lower_case=False,
skip_special_tokens=True),
device=0
)
sv_text = "EUROPEISKA GEMENSKAPERNAS RÅD HAR ANTAGIT DENNA FÖRORDNING med beaktande av Fördraget om upprättandet av Europeiska ekonomiska gemenskapen, särskilt artiklarna 43 och 100a i detta, med beaktande av kommissionens förslag(1), i samarbete med Europaparlamentet(2), med beaktande av Ekonomiska och sociala kommitténs yttrande(3), och med beaktande av följande: Det bör införas förbud mot användning av blybaserade kapsyler eller blybaserad folie i förslutningar på förpackningar som används då aromatiserade viner, aromatiserade vinbaserade drycker och aromatiserade drinkar baserade på vinprodukter släpps ut på marknaden i syfte att undvika risken för kontaminering, särskilt vid oavsiktlig kontakt med sådana produkter, samt risken för miljöförorening på grund av avfall som innehåller bly från kapsyler och folie av detta slag. Tillverkarna och användarna av kapsylerna och folien i fråga bör dock ges tid att anpassa sig genom att förbudet inte tillämpas förrän från och med den 1 januari 1993. Det är även nödvändigt att tillåta att produkter som före detta datum tappats på buteljer med blybaserade kapsyler eller blybaserad folie får säljas till dess att lagren är uttömda. Vissa definitioner av aromatiserade vinbaserade drycker bör anpassas så att större hänsyn tas till traditionella framställningsmetoder. Förordning (EEG) nr 1601/91(4) bör därför ändras. HÄRIGENOM FÖRESKRIVS FÖLJANDE. Artikel 1 Förordning (EEG) nr 1601/91 ändras på följande sätt: 1. Artikel 2.3 a första stycket skall ersättas med följande: %quot%a) Sangria: en dryck som framställs av vin - som smaksatts genom tillsats av naturliga extrakt eller essenser av citrusfrukt, - med eller utan saft av sådan frukt, - eventuellt: - med tillsats av kryddor, - sötat, - med tillsats av CO2, och med en slutlig alkoholstyrka på under 12 volymprocent.%quot% 2. Artikel 2.3 e skall ersättas med följande: %quot%e) Kalte Ente: Smaksatt vinbaserad dryck som framställs genom att vin, pärlande vin eller pärlande vin med tillsatt CO2 blandas med mousserande vin eller mousserande vin med tillsatt CO2 och tillsätts naturlig citronsubstans eller extrakt av detta som måste ge en tydligt framträdande smak. Slutprodukten måste innehålla minst 25 volymprocent mousserande vin eller mousserande vin med tillsatt CO2.%quot% 3. Följande punkt skall införas i artikel 8: %quot%4.a Från och med den 1 januari 1993 får buteljerade produkter som omfattas av denna förordning inte saluhållas eller släppas ut på marknaden i förpackningar med förslutningar som täckts med blybaserade kapsyler eller blybaserad folie. Dock får produkter som före detta datum tappats på flaskor med detta slag av kapsyler eller folie avyttras till dess att lagren tömts.%quot% Artikel 2 Denna förordning träder i kraft den tredje dagen efter det att den har offentliggjorts i Europeiska gemenskapernas officiella tidning. Denna förordning är till alla delar bindande och direkt tillämplig i alla medlemsstater. Utfärdad i Bryssel den 9 november 1992. På rådets vägnar D. HURD Ordförande (1) EGT nr C 69, 18.3.1992, s. 11. (2) EGT nr C 241, 21.9.1992, s. 97 och beslut av den 28 oktober 1992. (3) EGT nr C 169, 6.7.1992, s. 1. (4) EGT nr L 149, 14.6.1991, s. 1. "
pipeline([sv_text], max_length=512)
```
## Training data
The legal_t5_small_summ_sv model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html) dataset consisting of 19 Thousand texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 64). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for classification test dataset, achieves the following results:
Test results :
| Model | Rouge1 | Rouge2 | Rouge Lsum |
|:-----:|:-----:|:-----:|:-----:|
| legal_t5_small_summ_sv | 78.84|69.97 |77.59|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_summ_it | SEBIS | 2021-06-23T11:23:40Z | 19 | 2 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"summarization Italian model",
"dataset:jrc-acquis",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Italian
tags:
- summarization Italian model
datasets:
- jrc-acquis
widget:
- text: "LA COMMISSIONE DELLE COMUNITÀ EUROPEE, visto il trattato che istituisce la Comunità europea, visto il regolamento (CEE) n. 2082/92 del Consiglio, del 14 luglio 1992, relativo alle attestazioni di specificità dei prodotti agricoli ed alimentari(1), in particolare l'articolo 9, paragrafo 1, considerando quanto segue: (1) A norma dell'articolo 7 del regolamento (CEE) n. 2082/92, la Finlandia ha trasmesso alla Commissione una domanda di registrazione della denominazione %quot%Kalakukko%quot% quale attestazione di specificità. (2) La dicitura %quot%specialità tradizionale garantita%quot% può applicarsi soltanto a denominazioni figuranti nel summenzionato albo. (3) Nessuna dichiarazione di opposizione, ai sensi dell'articolo 8 del summenzionato regolamento, è stata trasmessa alla Commissione a seguito della pubblicazione nella Gazzetta ufficiale delle Comunità europee(2) della denominazione figurante nell'allegato del presente regolamento. (4) Di conseguenza, la denominazione di cui all'allegato può essere iscritta nell'albo delle attestazioni di specificità e beneficiare pertanto della protezione a livello comunitario quale specialità tradizionale garantita nella Comunità in virtù dell'articolo 13, paragrafo 2, del regolamento (CEE) n. 2082/92. (5) L'allegato del presente regolamento completa l'allegato del regolamento (CE) n. 2301/97 della Commissione(3), modificato da ultimo dal regolamento (CE) n. 688/2002(4), HA ADOTTATO IL PRESENTE REGOLAMENTO: Articolo 1 La denominazione di cui all'allegato del presente regolamento è aggiunta all'allegato del regolamento (CE) n. 2301/97 e iscritta nell'albo delle attestazioni di specificità, conformemente all'articolo 9, paragrafo 1, del regolamento (CEE) n. 2082/92. Tale denominazione è protetta ai sensi dell'articolo 13, paragrafo 2, del summenzionato regolamento. Articolo 2 Il presente regolamento entra in vigore il ventesimo giorno successivo alla pubblicazione nella Gazzetta ufficiale delle Comunità europee. Il presente regolamento è obbligatorio in tutti i suoi elementi e direttamente applicabile in ciascuno degli Stati membri. Fatto a Bruxelles, il 15 luglio 2002. Per la Commissione Franz Fischler Membro della Commissione (1) GU L 208 del 24.7.1992, pag. 9. (2) GU C 235 del 21.8.2001, pag. 12. (3) GU L 319 del 21.11.1997, pag. 8. (4) GU L 106 del 23.4.2002, pag. 7. ALLEGATO Prodotti della panetteria, della pasticceria, della confetteria o della biscotteria - Kalakukko "
---
# legal_t5_small_summ_it model
Model for Summarization of legal text written in Italian. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis.
## Model description
legal_t5_small_summ_it is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for summarization of legal texts written in Italian.
### How to use
Here is how to use this model to summarize legal text written in Italian in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_summ_it"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_summ_it", do_lower_case=False,
skip_special_tokens=True),
device=0
)
it_text = "LA COMMISSIONE DELLE COMUNITÀ EUROPEE, visto il trattato che istituisce la Comunità europea, visto il regolamento (CEE) n. 2082/92 del Consiglio, del 14 luglio 1992, relativo alle attestazioni di specificità dei prodotti agricoli ed alimentari(1), in particolare l'articolo 9, paragrafo 1, considerando quanto segue: (1) A norma dell'articolo 7 del regolamento (CEE) n. 2082/92, la Finlandia ha trasmesso alla Commissione una domanda di registrazione della denominazione %quot%Kalakukko%quot% quale attestazione di specificità. (2) La dicitura %quot%specialità tradizionale garantita%quot% può applicarsi soltanto a denominazioni figuranti nel summenzionato albo. (3) Nessuna dichiarazione di opposizione, ai sensi dell'articolo 8 del summenzionato regolamento, è stata trasmessa alla Commissione a seguito della pubblicazione nella Gazzetta ufficiale delle Comunità europee(2) della denominazione figurante nell'allegato del presente regolamento. (4) Di conseguenza, la denominazione di cui all'allegato può essere iscritta nell'albo delle attestazioni di specificità e beneficiare pertanto della protezione a livello comunitario quale specialità tradizionale garantita nella Comunità in virtù dell'articolo 13, paragrafo 2, del regolamento (CEE) n. 2082/92. (5) L'allegato del presente regolamento completa l'allegato del regolamento (CE) n. 2301/97 della Commissione(3), modificato da ultimo dal regolamento (CE) n. 688/2002(4), HA ADOTTATO IL PRESENTE REGOLAMENTO: Articolo 1 La denominazione di cui all'allegato del presente regolamento è aggiunta all'allegato del regolamento (CE) n. 2301/97 e iscritta nell'albo delle attestazioni di specificità, conformemente all'articolo 9, paragrafo 1, del regolamento (CEE) n. 2082/92. Tale denominazione è protetta ai sensi dell'articolo 13, paragrafo 2, del summenzionato regolamento. Articolo 2 Il presente regolamento entra in vigore il ventesimo giorno successivo alla pubblicazione nella Gazzetta ufficiale delle Comunità europee. Il presente regolamento è obbligatorio in tutti i suoi elementi e direttamente applicabile in ciascuno degli Stati membri. Fatto a Bruxelles, il 15 luglio 2002. Per la Commissione Franz Fischler Membro della Commissione (1) GU L 208 del 24.7.1992, pag. 9. (2) GU C 235 del 21.8.2001, pag. 12. (3) GU L 319 del 21.11.1997, pag. 8. (4) GU L 106 del 23.4.2002, pag. 7. ALLEGATO Prodotti della panetteria, della pasticceria, della confetteria o della biscotteria - Kalakukko "
pipeline([it_text], max_length=512)
```
## Training data
The legal_t5_small_summ_it model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html) dataset consisting of 22 Thousand texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 64). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for classification test dataset, achieves the following results:
Test results :
| Model | Rouge1 | Rouge2 | Rouge Lsum |
|:-----:|:-----:|:-----:|:-----:|
| legal_t5_small_summ_it | 75.07|65.53 |73.85|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_summ_fr | SEBIS | 2021-06-23T11:23:07Z | 4 | 1 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"summarization French model",
"dataset:jrc-acquis",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: French
tags:
- summarization French model
datasets:
- jrc-acquis
widget:
- text: "LA COMMISSION DES COMMUNAUTÉS EUROPÉENNES, vu le traité instituant la Communauté européenne, vu le règlement (CE) no 1784/2003 du Conseil du 29 septembre 2003 portant organisation commune des marchés dans le secteur des céréales [1], et notamment son article 13, paragraphe 3, vu le règlement (CE) no 1785/2003 du Conseil du 29 septembre 2003 portant organisation commune du marché du riz [2], et notamment son article 14, paragraphe 3, considérant ce qui suit: (1) Conformément à l'article 13, paragraphe 1, du règlement (CE) no 1784/2003 et à l'article 14, paragraphe 1, du règlement (CE) no 1785/2003, la différence entre les cours ou les prix sur le marché mondial des produits visés à l'article 1er de chacun de ces deux règlements et les prix dans la Communauté peut être couverte par une restitution à l'exportation. (2) Le règlement (CE) no 1043/2005 de la Commission du 30 juin 2005 portant application du règlement (CE) no 3448/93 du Conseil en ce qui concerne le système d’octroi des restitutions à l'exportation pour certains produits agricoles exportés sous forme de marchandises ne relevant pas de l'annexe I du traité ainsi que les critères de fixation de leurs montants [3] a spécifié ceux de ces produits pour lesquels il y a lieu de fixer un taux de restitution applicable lors de leur exportation sous forme de marchandises reprises, selon le cas, à l'annexe III du règlement (CE) no 1784/2003 ou à l'annexe IV du règlement (CE) no 1785/2003. (3) Conformément à l'article 14, paragraphe 1, du règlement (CE) no 1043/2005, le taux de la restitution par 100 kilogrammes de chacun des produits de base considérés doit être fixé chaque mois. (4) Les engagements pris en matière de restitutions pouvant être octroyées à l'exportation de produits agricoles incorporés dans des marchandises ne relevant pas de l'annexe I du traité peuvent être mis en péril par la fixation à l'avance de taux de restitution élevés. Il convient, dès lors, de prendre des mesures de sauvegarde dans ces situations sans empêcher pour autant la conclusion de contrats à long terme. La fixation d'un taux de restitution spécifique pour la fixation à l'avance des restitutions est une mesure permettant de rencontrer ces différents objectifs. (5) À la suite de l'arrangement entre la Communauté européenne et les États-Unis d'Amérique concernant les exportations de pâtes alimentaires de la Communauté aux États-Unis approuvé par la décision 87/482/CEE du Conseil [4], il est nécessaire de différencier la restitution pour les marchandises relevant des codes NC 19021100 et 190219 selon leur destination. (6) Conformément à l'article 15, paragraphes 2 et 3, du règlement (CE) no 1043/2005, il y a lieu de fixer un taux de restitution à l'exportation réduit, compte tenu du montant de la restitution à la production applicable, en vertu du règlement (CEE) no 1722/93 de la Commission [5], au produit de base mis en œuvre, valable au cours de la période présumée de fabrication des marchandises. (7) Les boissons spiritueuses sont considérées comme moins sensibles au prix des céréales mises en œuvre pour leur fabrication. Toutefois, le protocole 19 du traité d'adhésion du Royaume-Uni, de l'Irlande et du Danemark prévoit que des mesures nécessaires doivent être arrêtées afin de faciliter l'utilisation des céréales communautaires pour la fabrication de boissons spiritueuses obtenues à partir de céréales. Il convient donc d'adapter le taux de restitution applicable aux céréales exportées sous forme de boissons spiritueuses. (8) Le comité de gestion des céréales n'a pas émis d'avis dans le délai imparti par son président, A ARRÊTÉ LE PRÉSENT RÈGLEMENT: Article premier Les taux des restitutions applicables aux produits de base figurant à l'annexe I du règlement (CE) no 1043/2005 et à l'article 1er du règlement (CE) no 1784/2003 ou à l'article 1er du règlement (CE) no 1785/2003 modifié, qui sont exportés sous forme de marchandises reprises respectivement à l'annexe III du règlement (CE) no 1784/2003 ou à l'annexe IV du règlement (CE) no 1785/2003, sont fixés comme indiqué à l'annexe du présent règlement. Article 2 Le présent règlement entre en vigueur le 23 septembre 2005. Le présent règlement est obligatoire dans tous ses éléments et directement applicable dans tout État membre. Fait à Bruxelles, le 22 septembre 2005. Par la Commission Günter Verheugen Vice-président [1] JO L 270 du 21.10.2003, p. 78. [2] JO L 270 du 21.10.2003, p. 96. [3] JO L 172 du 5.7.2005, p. 24. [4] JO L 275 du 29.9.1987, p. 36. [5] JO L 159 du 1.7.1993, p. 112. Règlement modifié en dernier lieu par le règlement (CE) no 1584/2004 (JO L 280 du 31.8.2004, p. 11). -------------------------------------------------- ANNEXE Taux des restitutions applicables à compter du 23 septembre 2005 à certains produits des secteurs des céréales et du riz exportés sous forme de marchandises ne relevant pas de l'annexe I du traité [1] (en EUR/100 kg) | Code NC | Désignation des marchandises | Taux de la restitution par 100 kg du produit de base | En cas de fixation à l'avance des restitutions | Autres | 10011000 | Froment (blé) dur: | | | – en cas d'exportation de marchandises relevant des codes NC 190211 et 190219 vers les États-Unis d'Amérique | — | — | – dans les autres cas | — | — | 10019099 | Froment (blé) tendre et méteil: | | | – en cas d'exportation de marchandises relevant des codes NC 190211 et 190219 vers les États-Unis d'Amérique | — | — | – dans les autres cas: | | | – – en cas d'application de l'article 15, paragraphe 3, du règlement (CE) no 1043/2005 | — | — | – – en cas d'exportation de marchandises relevant du sous-chapitre 2208 | — | — | – – dans les autres cas | — | — | 10020000 | Seigle | — | — | 10030090 | Orge | | | – en cas d'exportation de marchandises relevant du sous-chapitre 2208 | — | — | – dans les autres cas | — | — | 10040000 | Avoine | — | — | 10059000 | Maïs, mis en œuvre sous forme de: | | | – amidon: | | | – – en cas d'application de l'article 15, paragraphe 3, du règlement (CE) no 1043/2005 | 2,994 | 3,150 | – – en cas d'exportation de marchandises relevant du sous-chapitre 2208 | 2,368 | 2,368 | – – dans les autres cas | 4,615 | 4,615 | – glucose, sirop de glucose, maltodextrine, sirop de maltodextrine des codes NC 17023051, 17023059, 17023091, 17023099, 17024090, 17029050, 17029075, 17029079, 21069055: | | | – – en cas d'application de l'article 15, paragraphe 3, du règlement (CE) no 1043/2005 | 1,840 | 1,996 | – – en cas d'exportation de marchandises relevant du sous-chapitre 2208 | 1,776 | 1,776 | – – dans les autres cas | 3,461 | 3,461 | – en cas d'exportation de marchandises relevant du sous-chapitre 2208 | 2,368 | 2,368 | – autres (y compris en l'état) | 4,615 | 4,615 | Fécule de pommes de terre du code NC 11081300 assimilée à un produit issu de la transformation du maïs: | | | – en cas d'application de l'article 15, paragraphe 3, du règlement (CE) no 1043/2005 | 2,435 | 2,585 | – en cas d'exportation de marchandises relevant du sous-chapitre 2208 | 2,368 | 2,368 | – dans les autres cas | 4,615 | 4,615 | ex100630 | Riz blanchi: | | | – à grains ronds | — | — | – à grains moyens | — | — | – à grains longs | — | — | 10064000 | Riz en brisures | — | — | 10070090 | Sorgho à grains (à l'excl. du sorgho à grains, hybride, destiné à l'ensemencement) | — | — | [1] Les taux prévus à la présente annexe ne s’appliquent pas avec effet au 1er octobre 2004 aux exportations vers la Bulgarie et avec effet au 1er février 2005 aux marchandises visées aux tableaux I et II du Protocole no 2 de l’Accord entre la Communauté économique européenne et la Confédération suisse du 22 juillet 1972 qui sont exportées vers la Confédération suisse ou la principauté de Liechtenstein. [2] En ce qui concerne les produits agricoles obtenus par transformation d’un produit de base et/ou de produits assimilés, les coefficients fixés à l’annexe V du règlement (CE) no 1043/2005 de la Commission s’appliquent. [3] La marchandise concernée relève du code NC 35051050. [4] Marchandises reprises à l'annexe III du règlement (CE) no 1784/2003 ou visées à l'article 2 du règlement (CEE) no 2825/93 (JO L 258 du 16.10.1993, p. 6). [5] Pour les sirops des codes NC 17023099, 17024090 et 17026090, obtenus par mélange de sirops de glucose et fructose, seul le sirop de glucose a droit à la restitution à l'exportation. -------------------------------------------------- "
---
# legal_t5_small_summ_fr model
Model for Summarization of legal text written in French. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis.
## Model description
legal_t5_small_summ_fr is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for summarization of legal texts written in French.
### How to use
Here is how to use this model to summarize legal text written in French in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_summ_fr"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_summ_fr", do_lower_case=False,
skip_special_tokens=True),
device=0
)
fr_text = "LA COMMISSION DES COMMUNAUTÉS EUROPÉENNES, vu le traité instituant la Communauté européenne, vu le règlement (CE) no 1784/2003 du Conseil du 29 septembre 2003 portant organisation commune des marchés dans le secteur des céréales [1], et notamment son article 13, paragraphe 3, vu le règlement (CE) no 1785/2003 du Conseil du 29 septembre 2003 portant organisation commune du marché du riz [2], et notamment son article 14, paragraphe 3, considérant ce qui suit: (1) Conformément à l'article 13, paragraphe 1, du règlement (CE) no 1784/2003 et à l'article 14, paragraphe 1, du règlement (CE) no 1785/2003, la différence entre les cours ou les prix sur le marché mondial des produits visés à l'article 1er de chacun de ces deux règlements et les prix dans la Communauté peut être couverte par une restitution à l'exportation. (2) Le règlement (CE) no 1043/2005 de la Commission du 30 juin 2005 portant application du règlement (CE) no 3448/93 du Conseil en ce qui concerne le système d’octroi des restitutions à l'exportation pour certains produits agricoles exportés sous forme de marchandises ne relevant pas de l'annexe I du traité ainsi que les critères de fixation de leurs montants [3] a spécifié ceux de ces produits pour lesquels il y a lieu de fixer un taux de restitution applicable lors de leur exportation sous forme de marchandises reprises, selon le cas, à l'annexe III du règlement (CE) no 1784/2003 ou à l'annexe IV du règlement (CE) no 1785/2003. (3) Conformément à l'article 14, paragraphe 1, du règlement (CE) no 1043/2005, le taux de la restitution par 100 kilogrammes de chacun des produits de base considérés doit être fixé chaque mois. (4) Les engagements pris en matière de restitutions pouvant être octroyées à l'exportation de produits agricoles incorporés dans des marchandises ne relevant pas de l'annexe I du traité peuvent être mis en péril par la fixation à l'avance de taux de restitution élevés. Il convient, dès lors, de prendre des mesures de sauvegarde dans ces situations sans empêcher pour autant la conclusion de contrats à long terme. La fixation d'un taux de restitution spécifique pour la fixation à l'avance des restitutions est une mesure permettant de rencontrer ces différents objectifs. (5) À la suite de l'arrangement entre la Communauté européenne et les États-Unis d'Amérique concernant les exportations de pâtes alimentaires de la Communauté aux États-Unis approuvé par la décision 87/482/CEE du Conseil [4], il est nécessaire de différencier la restitution pour les marchandises relevant des codes NC 19021100 et 190219 selon leur destination. (6) Conformément à l'article 15, paragraphes 2 et 3, du règlement (CE) no 1043/2005, il y a lieu de fixer un taux de restitution à l'exportation réduit, compte tenu du montant de la restitution à la production applicable, en vertu du règlement (CEE) no 1722/93 de la Commission [5], au produit de base mis en œuvre, valable au cours de la période présumée de fabrication des marchandises. (7) Les boissons spiritueuses sont considérées comme moins sensibles au prix des céréales mises en œuvre pour leur fabrication. Toutefois, le protocole 19 du traité d'adhésion du Royaume-Uni, de l'Irlande et du Danemark prévoit que des mesures nécessaires doivent être arrêtées afin de faciliter l'utilisation des céréales communautaires pour la fabrication de boissons spiritueuses obtenues à partir de céréales. Il convient donc d'adapter le taux de restitution applicable aux céréales exportées sous forme de boissons spiritueuses. (8) Le comité de gestion des céréales n'a pas émis d'avis dans le délai imparti par son président, A ARRÊTÉ LE PRÉSENT RÈGLEMENT: Article premier Les taux des restitutions applicables aux produits de base figurant à l'annexe I du règlement (CE) no 1043/2005 et à l'article 1er du règlement (CE) no 1784/2003 ou à l'article 1er du règlement (CE) no 1785/2003 modifié, qui sont exportés sous forme de marchandises reprises respectivement à l'annexe III du règlement (CE) no 1784/2003 ou à l'annexe IV du règlement (CE) no 1785/2003, sont fixés comme indiqué à l'annexe du présent règlement. Article 2 Le présent règlement entre en vigueur le 23 septembre 2005. Le présent règlement est obligatoire dans tous ses éléments et directement applicable dans tout État membre. Fait à Bruxelles, le 22 septembre 2005. Par la Commission Günter Verheugen Vice-président [1] JO L 270 du 21.10.2003, p. 78. [2] JO L 270 du 21.10.2003, p. 96. [3] JO L 172 du 5.7.2005, p. 24. [4] JO L 275 du 29.9.1987, p. 36. [5] JO L 159 du 1.7.1993, p. 112. Règlement modifié en dernier lieu par le règlement (CE) no 1584/2004 (JO L 280 du 31.8.2004, p. 11). -------------------------------------------------- ANNEXE Taux des restitutions applicables à compter du 23 septembre 2005 à certains produits des secteurs des céréales et du riz exportés sous forme de marchandises ne relevant pas de l'annexe I du traité [1] (en EUR/100 kg) | Code NC | Désignation des marchandises | Taux de la restitution par 100 kg du produit de base | En cas de fixation à l'avance des restitutions | Autres | 10011000 | Froment (blé) dur: | | | – en cas d'exportation de marchandises relevant des codes NC 190211 et 190219 vers les États-Unis d'Amérique | — | — | – dans les autres cas | — | — | 10019099 | Froment (blé) tendre et méteil: | | | – en cas d'exportation de marchandises relevant des codes NC 190211 et 190219 vers les États-Unis d'Amérique | — | — | – dans les autres cas: | | | – – en cas d'application de l'article 15, paragraphe 3, du règlement (CE) no 1043/2005 | — | — | – – en cas d'exportation de marchandises relevant du sous-chapitre 2208 | — | — | – – dans les autres cas | — | — | 10020000 | Seigle | — | — | 10030090 | Orge | | | – en cas d'exportation de marchandises relevant du sous-chapitre 2208 | — | — | – dans les autres cas | — | — | 10040000 | Avoine | — | — | 10059000 | Maïs, mis en œuvre sous forme de: | | | – amidon: | | | – – en cas d'application de l'article 15, paragraphe 3, du règlement (CE) no 1043/2005 | 2,994 | 3,150 | – – en cas d'exportation de marchandises relevant du sous-chapitre 2208 | 2,368 | 2,368 | – – dans les autres cas | 4,615 | 4,615 | – glucose, sirop de glucose, maltodextrine, sirop de maltodextrine des codes NC 17023051, 17023059, 17023091, 17023099, 17024090, 17029050, 17029075, 17029079, 21069055: | | | – – en cas d'application de l'article 15, paragraphe 3, du règlement (CE) no 1043/2005 | 1,840 | 1,996 | – – en cas d'exportation de marchandises relevant du sous-chapitre 2208 | 1,776 | 1,776 | – – dans les autres cas | 3,461 | 3,461 | – en cas d'exportation de marchandises relevant du sous-chapitre 2208 | 2,368 | 2,368 | – autres (y compris en l'état) | 4,615 | 4,615 | Fécule de pommes de terre du code NC 11081300 assimilée à un produit issu de la transformation du maïs: | | | – en cas d'application de l'article 15, paragraphe 3, du règlement (CE) no 1043/2005 | 2,435 | 2,585 | – en cas d'exportation de marchandises relevant du sous-chapitre 2208 | 2,368 | 2,368 | – dans les autres cas | 4,615 | 4,615 | ex100630 | Riz blanchi: | | | – à grains ronds | — | — | – à grains moyens | — | — | – à grains longs | — | — | 10064000 | Riz en brisures | — | — | 10070090 | Sorgho à grains (à l'excl. du sorgho à grains, hybride, destiné à l'ensemencement) | — | — | [1] Les taux prévus à la présente annexe ne s’appliquent pas avec effet au 1er octobre 2004 aux exportations vers la Bulgarie et avec effet au 1er février 2005 aux marchandises visées aux tableaux I et II du Protocole no 2 de l’Accord entre la Communauté économique européenne et la Confédération suisse du 22 juillet 1972 qui sont exportées vers la Confédération suisse ou la principauté de Liechtenstein. [2] En ce qui concerne les produits agricoles obtenus par transformation d’un produit de base et/ou de produits assimilés, les coefficients fixés à l’annexe V du règlement (CE) no 1043/2005 de la Commission s’appliquent. [3] La marchandise concernée relève du code NC 35051050. [4] Marchandises reprises à l'annexe III du règlement (CE) no 1784/2003 ou visées à l'article 2 du règlement (CEE) no 2825/93 (JO L 258 du 16.10.1993, p. 6). [5] Pour les sirops des codes NC 17023099, 17024090 et 17026090, obtenus par mélange de sirops de glucose et fructose, seul le sirop de glucose a droit à la restitution à l'exportation. -------------------------------------------------- "
pipeline([fr_text], max_length=512)
```
## Training data
The legal_t5_small_summ_fr model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html) dataset consisting of 23 Thousand texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 64). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for classification test dataset, achieves the following results:
Test results :
| Model | Rouge1 | Rouge2 | Rouge Lsum |
|:-----:|:-----:|:-----:|:-----:|
| legal_t5_small_summ_fr | 77.1|67.97 |75.74|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_summ_en | SEBIS | 2021-06-23T11:21:55Z | 37 | 1 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"summarization English model",
"dataset:jrc-acquis",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: English
tags:
- summarization English model
datasets:
- jrc-acquis
widget:
- text: >
THE COMMISSION OF THE EUROPEAN COMMUNITIES, Having regard to the Treaty establishing
the European Community, Having regard to Council Regulation (EC) No 1255/1999 of 17 May 1999
on the common organisation of the market in milk and milk products [1], and in particular Article 15 thereof,
Whereas: (1) Article 7(1) of Commission Regulation (EC) No 2799/1999 [2] fixes the amount of aid for
skimmed milk and skimmed-milk powder intended for animal feed taking into account the factors set out
in Article 11(2) of Regulation (EC) No 1255/1999. In view of the developments in the market price of
skimmed-milk powder, of the increase in the market prices for competing proteins, and of the reduction
of the supply of skimmed-milk powder, the amount of aid should be reduced. (2) Regulation (EC)
No 2799/1999 should therefore be amended accordingly. (3) The Management Committee for Milk and
Milk Products has not delivered an opinion within the time-limit set by its chairman,
HAS ADOPTED THIS REGULATION: Article 1 In Article 7 of Regulation (EC) No 2799/1999, paragraph 1 is replaced by the following: "1. Aid is fixed at: (a) EUR 1,62 per 100 kg of skimmed milk with a protein content of not less than 35,6 % of the non-fatty dry extract; (b) EUR 1,42 per 100 kg of skimmed milk with a protein content of not less than 31,4 % but less than 35,6 % of the non-fatty dry extract; (c) EUR 20,00 per 100 kg of skimmed-milk powder with a protein content of not less than 35,6 % of the non-fatty dry extract; (d) EUR 17,64 per 100 kg of skimmed-milk powder with a protein content of not less than 31,4 % but less than 35,6 % of the non-fatty dry extract." Article 2 This Regulation shall enter into force on the day following its publication in the Official Journal of the European Union. This Regulation shall be binding in its entirety and directly applicable in all Member States. Done at Brussels, 19 April 2006. For the Commission Mariann Fischer Boel Member of the Commission [1] OJ L 160, 26.6.1999, p. 48. Regulation as last amended by Regulation (EC) No 1913/2005 (OJ L 307, 25.11.2005, p. 2). [2] OJ L 340, 31.12.1999, p. 3.
Regulation as last amended by Regulation (EC) No 1194/2005 (OJ L 194, 26.7.2005, p. 7).
---
# legal_t5_small_summ_en model
Model for Summarization of legal text written in English. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis.
## Model description
legal_t5_small_summ_en is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for summarization of legal texts written in English.
### How to use
Here is how to use this model to summarize legal text written in English in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_summ_en"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_summ_en", do_lower_case=False,
skip_special_tokens=True),
device=0
)
en_text = "THE COMMISSION OF THE EUROPEAN COMMUNITIES, Having regard to the Treaty establishing the European Community, Having regard to Council Regulation (EC) No 1255/1999 of 17 May 1999 on the common organisation of the market in milk and milk products [1], and in particular Article 15 thereof, Whereas: (1) Article 7(1) of Commission Regulation (EC) No 2799/1999 [2] fixes the amount of aid for skimmed milk and skimmed-milk powder intended for animal feed taking into account the factors set out in Article 11(2) of Regulation (EC) No 1255/1999. In view of the developments in the market price of skimmed-milk powder, of the increase in the market prices for competing proteins, and of the reduction of the supply of skimmed-milk powder, the amount of aid should be reduced. (2) Regulation (EC) No 2799/1999 should therefore be amended accordingly. (3) The Management Committee for Milk and Milk Products has not delivered an opinion within the time-limit set by its chairman, HAS ADOPTED THIS REGULATION: Article 1 In Article 7 of Regulation (EC) No 2799/1999, paragraph 1 is replaced by the following: "1. Aid is fixed at: (a) EUR 1,62 per 100 kg of skimmed milk with a protein content of not less than 35,6 % of the non-fatty dry extract; (b) EUR 1,42 per 100 kg of skimmed milk with a protein content of not less than 31,4 % but less than 35,6 % of the non-fatty dry extract; (c) EUR 20,00 per 100 kg of skimmed-milk powder with a protein content of not less than 35,6 % of the non-fatty dry extract; (d) EUR 17,64 per 100 kg of skimmed-milk powder with a protein content of not less than 31,4 % but less than 35,6 % of the non-fatty dry extract." Article 2 This Regulation shall enter into force on the day following its publication in the Official Journal of the European Union. This Regulation shall be binding in its entirety and directly applicable in all Member States. Done at Brussels, 19 April 2006. For the Commission Mariann Fischer Boel Member of the Commission [1] OJ L 160, 26.6.1999, p. 48. Regulation as last amended by Regulation (EC) No 1913/2005 (OJ L 307, 25.11.2005, p. 2). [2] OJ L 340, 31.12.1999, p. 3. Regulation as last amended by Regulation (EC) No 1194/2005 (OJ L 194, 26.7.2005, p. 7). -------------------------------------------------- "
pipeline([en_text], max_length=512)
```
## Training data
The legal_t5_small_summ_en model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html) dataset consisting of 22 Thousand texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 64). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for classification test dataset, achieves the following results:
Test results :
| Model | Rouge1 | Rouge2 | Rouge Lsum |
|:-----:|:-----:|:-----:|:-----:|
| legal_t5_small_summ_en | 78.11|68.78 |77.0|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_summ_de | SEBIS | 2021-06-23T11:21:22Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"summarization Deustch model",
"dataset:jrc-acquis",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Deustch
tags:
- summarization Deustch model
datasets:
- jrc-acquis
widget:
- text: "(90/365/EWG) DER RAT DER EUROPÄISCHEN GEMEINSCHAFTEN - gestützt auf den Vertrag zur Gründung der Europäischen Wirtschaftsgemeinschaft, insbesondere auf Artikel 235, auf Vorschlag der Kommission (1), nach Stellungnahme des Europäischen Parlaments (2), nach Stellungnahme des Wirtschafts- und Sozialausschusses (3), in Erwägung nachstehender Gründe: Gemäß Artikel 3 Buchstabe c) des Vertrages umfasst die Tätigkeit der Gemeinschaft, nach Maßgabe des Vertrages, die Beseitigung der Hindernisse für den freien Personenverkehr zwischen den Mitgliedstaaten. Artikel 8a des Vertrages sieht vor, daß der Binnenmarkt bis zum 31. Dezember 1992 zu verwirklichen ist. Der Binnenmarkt umfasst einen Raum ohne Binnengrenzen, in dem der freie Verkehr von Waren, Personen, Dienstleistungen und Kapital gemäß den Bestimmungen des Vertrages gewährleistet ist. Die Artikel 48 und 52 des Vertrages sehen die Freizuegigkeit der Arbeitnehmer und selbständig Erwerbstätigen vor, was ein Recht auf Aufenthalt in dem Mitgliedstaat beinhaltet, in dem sie ihr Berufsleben verbringen. Es empfiehlt sich, dieses Aufenthaltsrecht auch Personen zu gewähren, die aus dem Erwerbsleben ausgeschieden sind, auch wenn sie während ihres Berufslebens von dem Recht auf Freizuegigkeit keinen Gebrauch gemacht haben. Die Aufenthaltsberechtigten dürfen die öffentlichen Finanzen des Aufnahmemitgliedstaates nicht über Gebühr belasten. Nach Artikel 10 der Verordnung (EWG) Nr. 1408/71 (4) in der Fassung der Verordnung (EWG) Nr. 1390/81 (5) haben die Empfänger von Geldleistungen bei Invalidität und Alter und die Bezieher von Renten bei Arbeitsunfällen oder Berufskrankheiten auch dann weiterhin Anspruch auf diese Leistungen und Renten, wenn sie im Gebiet eines anderen Mitgliedstaates als des Staates wohnen, auf dessen Gebiet der zur Zahlung verpflichtete Träger seinen Sitz hat. Die Ausübung des Aufenthaltsrechts wird erst dann eine reale Möglichkeit, wenn es auch den Familienangehörigen zugestanden wird. Für die von dieser Richtlinie Begünstigten sollte eine Verwaltungsregelung entsprechend der insbesondere in der Richtlinie 68/360/EWG (6) und in der Richtlinie 64/221/EWG (7) vorgesehenen Regelung gelten. Der Vertrag enthält Befugnisse für den Erlaß der vorliegenden Richtlinie nur in Artikel 235 - HAT FOLGENDE RICHTLINIE ERLASSEN: Artikel 1 (1) Die Mitgliedstaaten gewähren den Angehörigen der Mitgliedstaaten, die in der Gemeinschaft eine Tätigkeit als Arbeitnehmer oder als Selbständige ausgeuebt haben, sowie deren Familienangehörigen nach der Definition von Absatz 2 unter der Bedingung das Aufenthaltsrecht, daß sie eine Invaliditäts-, Vorruhestands- oder Altersrente oder eine Rente wegen Arbeitsunfalls oder Berufskrankheit in einer solchen Höhe beziehen, daß sie während ihres Aufenthalts nicht die Sozialhilfe des Aufnahmemitgliedstaats in Anspruch nehmen müssen, und einen Krankenversicherungsschutz genießen, der im Aufnahmemitgliedstaat alle Risiken abdeckt. Die Existenzmittel des Antragstellers gelten als ausreichend, wenn sie einen Betrag übersteigen, unterhalb dessen der Aufnahmemitgliedstaat seinen Staatsangehörigen aufgrund der persönlichen Situation des Antragstellers und gegebenenfalls der Situation der nach Absatz 2 aufgenommenen Personen Sozialhilfe gewähren kann. Ist Unterabsatz 2 in einem Mitgliedstaat nicht anwendbar, so gelten die Existenzmittel des Antragstellers als ausreichend, wenn sie den Betrag der Grundrente der Sozialversicherung übersteigen, die der Aufnahmemitgliedstaat zahlt. (2) Bei dem Aufenthaltsberechtigten dürfen folgende Personen ungeachtet ihrer Staatsangehörigkeit in einem anderen Mitgliedstaat Wohnung nehmen: a) sein Ehegatte sowie die Verwandten in absteigender Linie, denen Unterhalt gewährt wird; b) seine Verwandten und die Verwandten seines Ehegatten in aufsteigender Linie, denen er Unterhalt gewährt. Artikel 2 (1) Zum Nachweis des Aufenthaltsrechts wird eine Bescheinigung, die »Aufenthaltserlaubnis für Staatsangehörige eines EWG-Mitgliedstaates%quot%, erteilt, deren Gültigkeit auf fünf Jahre mit Verlängerungsmöglichkeit begrenzt werden kann. Die Mitgliedstaaten können jedoch die Erneuerung der Aufenthaltserlaubnis nach den ersten zwei Aufenthaltsjahren verlangen, wenn sie dies für erforderlich halten. Einem Familienmitglied, das nicht die Staatsangehörigkeit eines Mitgliedstaats besitzt, wird ein Aufenthaltsdokument mit der gleichen Gültigkeitsdauer ausgestellt wie dem Staatsangehörigen, von dem es seine Rechte herleitet. Für die Erteilung der Aufenthaltserlaubnis oder des Aufenthaltsdokuments darf der Mitgliedstaat vom Antragsteller nur die Vorlage eines gültigen Personalausweises bzw. Reisepasses sowie den Nachweis verlangen, daß er die Voraussetzungen des Artikels 1 erfuellt. (2) Die Artikel 2 und 3, Artikel 6 Absatz 1 Buchstabe a) und Absatz 2 sowie Artikel 9 der Richtlinie 68/360/EWG finden auf die von dieser Richtlinie Begünstigten entsprechende Anwendung. Der Ehegatte eines Staatsangehörigen eines Mitgliedstaats, der im Hoheitsgebiet eines Mitgliedstaats aufenthaltsberechtigt ist, sowie die Kinder dieses Staatsangehörigen, denen er Unterhalt gewährt, haben, auch wenn sie die Staatsangehörigkeit eines Mitgliedstaats nicht besitzen, das Recht, im gesamten Hoheitsgebiet dieses Mitgliedstaats jedwede Tätigkeit im Lohn- oder Gehaltsverhältnis oder jedwede selbständige Erwerbstätigkeit auszuüben. Die Mitgliedstaaten dürfen nur aus Gründen der öffentlichen Ordnung, der öffentlichen Sicherheit oder der Volksgesundheit von den Bestimmungen dieser Richtlinie abweichen. In diesem Fall findet die Richtlinie 64/221/EWG Anwendung. (3) Die vorliegende Richtlinie berührt nicht die geltenden Rechtsvorschriften für den Erwerb von Zweitwohnungen. Artikel 3 Das Aufenthaltsrecht besteht, solange die Berechtigten die Bedingungen des Artikels 1 erfuellen. Artikel 4 Die Kommission arbeitet spätestens drei Jahre nach dem Beginn der Anwendung dieser Richtlinie und anschließend alle drei Jahre einen Bericht über ihre Anwendung aus und legt ihn dem Europäischen Parlament und dem Rat vor. Artikel 5 Die Mitgliedstaaten setzen die erforderlichen Rechts- und Verwaltungsvorschriften in Kraft, um dieser Richtlinie bis spätestens 30. Juni 1992 nachzukommen. Sie setzen die Kommission unverzueglich davon in Kenntnis. Artikel 6 Diese Richtlinie ist an die Mitgliedstaaten gerichtet. Geschehen zu Luxemburg am 28. Juni 1990. Im Namen des Rates Der Präsident M. GEOGHEGAN-QUINN (1) ABl. Nr. C 191 vom 28. 7. 1989, S. 3 und ABl. Nr. C 26 vom 3. 2. 1990, S. 19. (2) Stellungnahme vom 13. Juni 1990 (noch nicht im Amtsblatt veröffentlicht). (3) ABl. Nr. C 329 vom 30. 12. 1989, S. 25. (4) ABl. Nr. L 149 vom 5. 7. 1971, S. 2. (5) ABl. Nr. L 143 vom 29. 5. 1981, S. 1. (6) ABl. Nr. L 257 vom 19. 10. 1968, S. 13. (7) ABl. Nr. 56 vom 4. 4. 1964, S. 850/64. "
---
# legal_t5_small_summ_de model
Model for Summarization of legal text written in Deustch. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis.
## Model description
legal_t5_small_summ_de is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for summarization of legal texts written in Deustch.
### How to use
Here is how to use this model to summarize legal text written in Deustch in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_summ_de"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_summ_de", do_lower_case=False,
skip_special_tokens=True),
device=0
)
de_text = "(90/365/EWG) DER RAT DER EUROPÄISCHEN GEMEINSCHAFTEN - gestützt auf den Vertrag zur Gründung der Europäischen Wirtschaftsgemeinschaft, insbesondere auf Artikel 235, auf Vorschlag der Kommission (1), nach Stellungnahme des Europäischen Parlaments (2), nach Stellungnahme des Wirtschafts- und Sozialausschusses (3), in Erwägung nachstehender Gründe: Gemäß Artikel 3 Buchstabe c) des Vertrages umfasst die Tätigkeit der Gemeinschaft, nach Maßgabe des Vertrages, die Beseitigung der Hindernisse für den freien Personenverkehr zwischen den Mitgliedstaaten. Artikel 8a des Vertrages sieht vor, daß der Binnenmarkt bis zum 31. Dezember 1992 zu verwirklichen ist. Der Binnenmarkt umfasst einen Raum ohne Binnengrenzen, in dem der freie Verkehr von Waren, Personen, Dienstleistungen und Kapital gemäß den Bestimmungen des Vertrages gewährleistet ist. Die Artikel 48 und 52 des Vertrages sehen die Freizuegigkeit der Arbeitnehmer und selbständig Erwerbstätigen vor, was ein Recht auf Aufenthalt in dem Mitgliedstaat beinhaltet, in dem sie ihr Berufsleben verbringen. Es empfiehlt sich, dieses Aufenthaltsrecht auch Personen zu gewähren, die aus dem Erwerbsleben ausgeschieden sind, auch wenn sie während ihres Berufslebens von dem Recht auf Freizuegigkeit keinen Gebrauch gemacht haben. Die Aufenthaltsberechtigten dürfen die öffentlichen Finanzen des Aufnahmemitgliedstaates nicht über Gebühr belasten. Nach Artikel 10 der Verordnung (EWG) Nr. 1408/71 (4) in der Fassung der Verordnung (EWG) Nr. 1390/81 (5) haben die Empfänger von Geldleistungen bei Invalidität und Alter und die Bezieher von Renten bei Arbeitsunfällen oder Berufskrankheiten auch dann weiterhin Anspruch auf diese Leistungen und Renten, wenn sie im Gebiet eines anderen Mitgliedstaates als des Staates wohnen, auf dessen Gebiet der zur Zahlung verpflichtete Träger seinen Sitz hat. Die Ausübung des Aufenthaltsrechts wird erst dann eine reale Möglichkeit, wenn es auch den Familienangehörigen zugestanden wird. Für die von dieser Richtlinie Begünstigten sollte eine Verwaltungsregelung entsprechend der insbesondere in der Richtlinie 68/360/EWG (6) und in der Richtlinie 64/221/EWG (7) vorgesehenen Regelung gelten. Der Vertrag enthält Befugnisse für den Erlaß der vorliegenden Richtlinie nur in Artikel 235 - HAT FOLGENDE RICHTLINIE ERLASSEN: Artikel 1 (1) Die Mitgliedstaaten gewähren den Angehörigen der Mitgliedstaaten, die in der Gemeinschaft eine Tätigkeit als Arbeitnehmer oder als Selbständige ausgeuebt haben, sowie deren Familienangehörigen nach der Definition von Absatz 2 unter der Bedingung das Aufenthaltsrecht, daß sie eine Invaliditäts-, Vorruhestands- oder Altersrente oder eine Rente wegen Arbeitsunfalls oder Berufskrankheit in einer solchen Höhe beziehen, daß sie während ihres Aufenthalts nicht die Sozialhilfe des Aufnahmemitgliedstaats in Anspruch nehmen müssen, und einen Krankenversicherungsschutz genießen, der im Aufnahmemitgliedstaat alle Risiken abdeckt. Die Existenzmittel des Antragstellers gelten als ausreichend, wenn sie einen Betrag übersteigen, unterhalb dessen der Aufnahmemitgliedstaat seinen Staatsangehörigen aufgrund der persönlichen Situation des Antragstellers und gegebenenfalls der Situation der nach Absatz 2 aufgenommenen Personen Sozialhilfe gewähren kann. Ist Unterabsatz 2 in einem Mitgliedstaat nicht anwendbar, so gelten die Existenzmittel des Antragstellers als ausreichend, wenn sie den Betrag der Grundrente der Sozialversicherung übersteigen, die der Aufnahmemitgliedstaat zahlt. (2) Bei dem Aufenthaltsberechtigten dürfen folgende Personen ungeachtet ihrer Staatsangehörigkeit in einem anderen Mitgliedstaat Wohnung nehmen: a) sein Ehegatte sowie die Verwandten in absteigender Linie, denen Unterhalt gewährt wird; b) seine Verwandten und die Verwandten seines Ehegatten in aufsteigender Linie, denen er Unterhalt gewährt. Artikel 2 (1) Zum Nachweis des Aufenthaltsrechts wird eine Bescheinigung, die »Aufenthaltserlaubnis für Staatsangehörige eines EWG-Mitgliedstaates%quot%, erteilt, deren Gültigkeit auf fünf Jahre mit Verlängerungsmöglichkeit begrenzt werden kann. Die Mitgliedstaaten können jedoch die Erneuerung der Aufenthaltserlaubnis nach den ersten zwei Aufenthaltsjahren verlangen, wenn sie dies für erforderlich halten. Einem Familienmitglied, das nicht die Staatsangehörigkeit eines Mitgliedstaats besitzt, wird ein Aufenthaltsdokument mit der gleichen Gültigkeitsdauer ausgestellt wie dem Staatsangehörigen, von dem es seine Rechte herleitet. Für die Erteilung der Aufenthaltserlaubnis oder des Aufenthaltsdokuments darf der Mitgliedstaat vom Antragsteller nur die Vorlage eines gültigen Personalausweises bzw. Reisepasses sowie den Nachweis verlangen, daß er die Voraussetzungen des Artikels 1 erfuellt. (2) Die Artikel 2 und 3, Artikel 6 Absatz 1 Buchstabe a) und Absatz 2 sowie Artikel 9 der Richtlinie 68/360/EWG finden auf die von dieser Richtlinie Begünstigten entsprechende Anwendung. Der Ehegatte eines Staatsangehörigen eines Mitgliedstaats, der im Hoheitsgebiet eines Mitgliedstaats aufenthaltsberechtigt ist, sowie die Kinder dieses Staatsangehörigen, denen er Unterhalt gewährt, haben, auch wenn sie die Staatsangehörigkeit eines Mitgliedstaats nicht besitzen, das Recht, im gesamten Hoheitsgebiet dieses Mitgliedstaats jedwede Tätigkeit im Lohn- oder Gehaltsverhältnis oder jedwede selbständige Erwerbstätigkeit auszuüben. Die Mitgliedstaaten dürfen nur aus Gründen der öffentlichen Ordnung, der öffentlichen Sicherheit oder der Volksgesundheit von den Bestimmungen dieser Richtlinie abweichen. In diesem Fall findet die Richtlinie 64/221/EWG Anwendung. (3) Die vorliegende Richtlinie berührt nicht die geltenden Rechtsvorschriften für den Erwerb von Zweitwohnungen. Artikel 3 Das Aufenthaltsrecht besteht, solange die Berechtigten die Bedingungen des Artikels 1 erfuellen. Artikel 4 Die Kommission arbeitet spätestens drei Jahre nach dem Beginn der Anwendung dieser Richtlinie und anschließend alle drei Jahre einen Bericht über ihre Anwendung aus und legt ihn dem Europäischen Parlament und dem Rat vor. Artikel 5 Die Mitgliedstaaten setzen die erforderlichen Rechts- und Verwaltungsvorschriften in Kraft, um dieser Richtlinie bis spätestens 30. Juni 1992 nachzukommen. Sie setzen die Kommission unverzueglich davon in Kenntnis. Artikel 6 Diese Richtlinie ist an die Mitgliedstaaten gerichtet. Geschehen zu Luxemburg am 28. Juni 1990. Im Namen des Rates Der Präsident M. GEOGHEGAN-QUINN (1) ABl. Nr. C 191 vom 28. 7. 1989, S. 3 und ABl. Nr. C 26 vom 3. 2. 1990, S. 19. (2) Stellungnahme vom 13. Juni 1990 (noch nicht im Amtsblatt veröffentlicht). (3) ABl. Nr. C 329 vom 30. 12. 1989, S. 25. (4) ABl. Nr. L 149 vom 5. 7. 1971, S. 2. (5) ABl. Nr. L 143 vom 29. 5. 1981, S. 1. (6) ABl. Nr. L 257 vom 19. 10. 1968, S. 13. (7) ABl. Nr. 56 vom 4. 4. 1964, S. 850/64. "
pipeline([de_text], max_length=512)
```
## Training data
The legal_t5_small_summ_de model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html) dataset consisting of 23 Thousand texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 64). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for classification test dataset, achieves the following results:
Test results :
| Model | Rouge1 | Rouge2 | Rouge Lsum |
|:-----:|:-----:|:-----:|:-----:|
| legal_t5_small_summ_de | 78.03|68.84 |76.95|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_summ_cs | SEBIS | 2021-06-23T11:20:42Z | 6 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"summarization Cszech model",
"dataset:jrc-acquis",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech
tags:
- summarization Cszech model
datasets:
- jrc-acquis
widget:
- text: "(2006/C 67/15) (Text s významem pro EHP) Dne 10. března 2006 se Komise rozhodla nevznést námitky proti výše uvedenému spojení a prohlásit ho za slučitelné se společným trhem. Toto rozhodnutí je založeno na čl. 6 odst. 1 písm. b) nařízení Rady (ES) č. 139/2004. Celý text rozhodnutí je přístupný pouze v angličtině a bude uveřejněn poté, co bude zbaven obchodního tajemství, které může případně obsahovat. Text bude dosažitelný: - na webové stránce Europa – hospodářská soutěž (http://europa.eu.int/comm/competition/mergers/cases/). Tato webová stránka umožňuje vyhledat jednotlivá rozhodnutí o spojení, a to včetně společnosti, čísla případu, data a indexu odvětví hospodářství. - v elektronické podobě na webové stránce EUR-Lex, pod dokumentem č. 32006M4093. EUR-Lex umožňuje přístup k Evropskému právu přes Internet. (http://europa.eu.int/eur-lex/lex) -------------------------------------------------- "
---
# legal_t5_small_summ_cs model
Model for Summarization of legal text written in Cszech. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis.
## Model description
legal_t5_small_summ_cs is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for summarization of legal texts written in Cszech.
### How to use
Here is how to use this model to summarize legal text written in Cszech in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_summ_cs"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_summ_cs", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "(2006/C 67/15) (Text s významem pro EHP) Dne 10. března 2006 se Komise rozhodla nevznést námitky proti výše uvedenému spojení a prohlásit ho za slučitelné se společným trhem. Toto rozhodnutí je založeno na čl. 6 odst. 1 písm. b) nařízení Rady (ES) č. 139/2004. Celý text rozhodnutí je přístupný pouze v angličtině a bude uveřejněn poté, co bude zbaven obchodního tajemství, které může případně obsahovat. Text bude dosažitelný: - na webové stránce Europa – hospodářská soutěž (http://europa.eu.int/comm/competition/mergers/cases/). Tato webová stránka umožňuje vyhledat jednotlivá rozhodnutí o spojení, a to včetně společnosti, čísla případu, data a indexu odvětví hospodářství. - v elektronické podobě na webové stránce EUR-Lex, pod dokumentem č. 32006M4093. EUR-Lex umožňuje přístup k Evropskému právu přes Internet. (http://europa.eu.int/eur-lex/lex) -------------------------------------------------- "
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_summ_cs model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html) dataset consisting of 18 Thousand texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 64). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for classification test dataset, achieves the following results:
Test results :
| Model | Rouge1 | Rouge2 | Rouge Lsum |
|:-----:|:-----:|:-----:|:-----:|
| legal_t5_small_summ_cs | 75.86|65.82 |74.95|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
huggingtweets/newcastle | huggingtweets | 2021-06-23T11:20:39Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"gpt2",
"text-generation",
"huggingtweets",
"en",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text-generation | 2022-03-02T23:29:05Z | ---
language: en
thumbnail: https://www.huggingtweets.com/newcastle/1624447235109/predictions.png
tags:
- huggingtweets
widget:
- text: "My dream is"
---
<div class="inline-flex flex-col" style="line-height: 1.5;">
<div class="flex">
<div
style="display:inherit; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://pbs.twimg.com/profile_images/378800000825693392/0c26d155e1abb8252f569491678b6ec7_400x400.jpeg')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
<div
style="display:none; margin-left: 4px; margin-right: 4px; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('')">
</div>
</div>
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 AI BOT 🤖</div>
<div style="text-align: center; font-size: 16px; font-weight: 800">Newcastle Brown Ale</div>
<div style="text-align: center; font-size: 14px;">@newcastle</div>
</div>
I was made with [huggingtweets](https://github.com/borisdayma/huggingtweets).
Create your own bot based on your favorite user with [the demo](https://colab.research.google.com/github/borisdayma/huggingtweets/blob/master/huggingtweets-demo.ipynb)!
## How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the [W&B report](https://wandb.ai/wandb/huggingtweets/reports/HuggingTweets-Train-a-Model-to-Generate-Tweets--VmlldzoxMTY5MjI).
## Training data
The model was trained on tweets from Newcastle Brown Ale.
| Data | Newcastle Brown Ale |
| --- | --- |
| Tweets downloaded | 3198 |
| Retweets | 21 |
| Short tweets | 27 |
| Tweets kept | 3150 |
[Explore the data](https://wandb.ai/wandb/huggingtweets/runs/1m1ygycf/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
## Training procedure
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on @newcastle's tweets.
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/wandb/huggingtweets/runs/2q9cnfvw) for full transparency and reproducibility.
At the end of training, [the final model](https://wandb.ai/wandb/huggingtweets/runs/2q9cnfvw/artifacts) is logged and versioned.
## How to use
You can use this model directly with a pipeline for text generation:
```python
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/newcastle')
generator("My dream is", num_return_sequences=5)
```
## Limitations and bias
The model suffers from [the same limitations and bias as GPT-2](https://huggingface.co/gpt2#limitations-and-bias).
In addition, the data present in the user's tweets further affects the text generated by the model.
## About
*Built by Boris Dayma*
[](https://twitter.com/intent/follow?screen_name=borisdayma)
For more details, visit the project repository.
[](https://github.com/borisdayma/huggingtweets)
|
SEBIS/legal_t5_small_multitask_sv_it | SEBIS | 2021-06-23T11:20:05Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Swedish Italian model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Swedish Italian
tags:
- translation Swedish Italian model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "De nationella tillsynsmyndigheterna får använda"
---
# legal_t5_small_multitask_sv_it model
Model on translating legal text from Swedish to Italian. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_sv_it model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Swedish to Italian.
### How to use
Here is how to use this model to translate legal text from Swedish to Italian in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_sv_it"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_sv_it", do_lower_case=False,
skip_special_tokens=True),
device=0
)
sv_text = "De nationella tillsynsmyndigheterna får använda"
pipeline([sv_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_sv_it model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 8 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_sv_it | 44.242|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_sv_fr | SEBIS | 2021-06-23T11:19:29Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Swedish French model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Swedish French
tags:
- translation Swedish French model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Europaparlamentet understryker att det stora antalet kvinnor och barn bland flyktingar och internt fördrivna som registrerats av internationella organ som resultat av väpnade konflikter och inbördeskrig är mycket oroväckande."
---
# legal_t5_small_multitask_sv_fr model
Model on translating legal text from Swedish to French. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_sv_fr model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Swedish to French.
### How to use
Here is how to use this model to translate legal text from Swedish to French in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_sv_fr"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_sv_fr", do_lower_case=False,
skip_special_tokens=True),
device=0
)
sv_text = "Europaparlamentet understryker att det stora antalet kvinnor och barn bland flyktingar och internt fördrivna som registrerats av internationella organ som resultat av väpnade konflikter och inbördeskrig är mycket oroväckande."
pipeline([sv_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_sv_fr model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 8 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_sv_fr | 45.790|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_sv_en | SEBIS | 2021-06-23T11:18:13Z | 6 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Swedish English model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Swedish English
tags:
- translation Swedish English model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "inlämnat av följande ledamöter:"
---
# legal_t5_small_multitask_sv_en model
Model on translating legal text from Swedish to English. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_sv_en model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Swedish to English.
### How to use
Here is how to use this model to translate legal text from Swedish to English in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_sv_en"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_sv_en", do_lower_case=False,
skip_special_tokens=True),
device=0
)
sv_text = "inlämnat av följande ledamöter:"
pipeline([sv_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_sv_en model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 9 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_sv_en | 36.195|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_it_sv | SEBIS | 2021-06-23T11:16:13Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Italian Swedish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Italian Swedish
tags:
- translation Italian Swedish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Può il Commissario responsabile comunicare al Parlamento in che modo la DG Ricerca garantirà che l’Europa possa svolgere un ruolo di primo piano in questo sforzo globale di ricerca sul diabete?"
---
# legal_t5_small_multitask_it_sv model
Model on translating legal text from Italian to Swedish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_it_sv model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Italian to Swedish.
### How to use
Here is how to use this model to translate legal text from Italian to Swedish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_it_sv"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_it_sv", do_lower_case=False,
skip_special_tokens=True),
device=0
)
it_text = "Può il Commissario responsabile comunicare al Parlamento in che modo la DG Ricerca garantirà che l’Europa possa svolgere un ruolo di primo piano in questo sforzo globale di ricerca sul diabete?"
pipeline([it_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_it_sv model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 8 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_it_sv | 41.523|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_it_es | SEBIS | 2021-06-23T11:14:49Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Italian Spanish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Italian Spanish
tags:
- translation Italian Spanish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Interrogazione con richiesta di risposta scritta E-005808/2011"
---
# legal_t5_small_multitask_it_es model
Model on translating legal text from Italian to Spanish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_it_es model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Italian to Spanish.
### How to use
Here is how to use this model to translate legal text from Italian to Spanish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_it_es"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_it_es", do_lower_case=False,
skip_special_tokens=True),
device=0
)
it_text = "Interrogazione con richiesta di risposta scritta E-005808/2011"
pipeline([it_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_it_es model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 9 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_it_es | 36.980|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_it_cs | SEBIS | 2021-06-23T11:12:39Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Italian Cszech model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Italian Cszech
tags:
- translation Italian Cszech model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Per mobilitare il Fondo, la Commissione ha presentato all'autorità di bilancio una richiesta di storno per un importo complessivo di 667.823 EUR dalla riserva FEG (40 02 43) in stanziamenti d'impegno verso la linea di bilancio FEG."
---
# legal_t5_small_multitask_it_cs model
Model on translating legal text from Italian to Cszech. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_it_cs model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Italian to Cszech.
### How to use
Here is how to use this model to translate legal text from Italian to Cszech in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_it_cs"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_it_cs", do_lower_case=False,
skip_special_tokens=True),
device=0
)
it_text = "Per mobilitare il Fondo, la Commissione ha presentato all'autorità di bilancio una richiesta di storno per un importo complessivo di 667.823 EUR dalla riserva FEG (40 02 43) in stanziamenti d'impegno verso la linea di bilancio FEG."
pipeline([it_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_it_cs model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_it_cs | 37.935|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_fr_sv | SEBIS | 2021-06-23T11:12:04Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation French Swedish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: French Swedish
tags:
- translation French Swedish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "**I Procédure de coopération (première lecture)"
---
# legal_t5_small_multitask_fr_sv model
Model on translating legal text from French to Swedish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_fr_sv model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from French to Swedish.
### How to use
Here is how to use this model to translate legal text from French to Swedish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_fr_sv"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_fr_sv", do_lower_case=False,
skip_special_tokens=True),
device=0
)
fr_text = "**I Procédure de coopération (première lecture)"
pipeline([fr_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_fr_sv model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 8 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_fr_sv | 39.947|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_fr_it | SEBIS | 2021-06-23T11:11:18Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation French Italian model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: French Italian
tags:
- translation French Italian model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Situation humanitaire au Soudan"
---
# legal_t5_small_multitask_fr_it model
Model on translating legal text from French to Italian. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_fr_it model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from French to Italian.
### How to use
Here is how to use this model to translate legal text from French to Italian in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_fr_it"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_fr_it", do_lower_case=False,
skip_special_tokens=True),
device=0
)
fr_text = "Situation humanitaire au Soudan"
pipeline([fr_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_fr_it model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 9 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_fr_it | 41.140|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_fr_en | SEBIS | 2021-06-23T11:10:07Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation French English model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: French English
tags:
- translation French English model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Raül Romeva i Rueda (Verts/ALE)"
---
# legal_t5_small_multitask_fr_en model
Model on translating legal text from French to English. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_fr_en model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from French to English.
### How to use
Here is how to use this model to translate legal text from French to English in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_fr_en"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_fr_en", do_lower_case=False,
skip_special_tokens=True),
device=0
)
fr_text = "Raül Romeva i Rueda (Verts/ALE)"
pipeline([fr_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_fr_en model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 9 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_fr_en | 39.123|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_es_en | SEBIS | 2021-06-23T11:03:37Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Spanish English model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Spanish English
tags:
- translation Spanish English model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "PPE-DE: 6', PSE: 6', ALDE: 5', Verts/ALE: 4', GUE/NGL: 4', IND/DEM:4', UEN: 4', NI: 4'"
---
# legal_t5_small_multitask_es_en model
Model on translating legal text from Spanish to English. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_es_en model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Spanish to English.
### How to use
Here is how to use this model to translate legal text from Spanish to English in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_es_en"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_es_en", do_lower_case=False,
skip_special_tokens=True),
device=0
)
es_text = "PPE-DE: 6', PSE: 6', ALDE: 5', Verts/ALE: 4', GUE/NGL: 4', IND/DEM:4', UEN: 4', NI: 4'"
pipeline([es_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_es_en model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 9 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_es_en | 36.607|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_es_de | SEBIS | 2021-06-23T11:02:08Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Spanish Deustch model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Spanish Deustch
tags:
- translation Spanish Deustch model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Estudios y publicaciones realizados por el Parlamento Europeo"
---
# legal_t5_small_multitask_es_de model
Model on translating legal text from Spanish to Deustch. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_es_de model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Spanish to Deustch.
### How to use
Here is how to use this model to translate legal text from Spanish to Deustch in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_es_de"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_es_de", do_lower_case=False,
skip_special_tokens=True),
device=0
)
es_text = "Estudios y publicaciones realizados por el Parlamento Europeo"
pipeline([es_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_es_de model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 8 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_es_de | 41.196|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_es_cs | SEBIS | 2021-06-23T11:01:33Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Spanish Cszech model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Spanish Cszech
tags:
- translation Spanish Cszech model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "La política pesquera supone que se tenga en cuenta un gran número de dimensiones – social, medioambiental, económica – lo que exige un enfoque integrado y equilibrado, incompatible con una visión que los sobrestima, en particular, mediante una definición a priori de cualquier jerarquía de prioridades."
---
# legal_t5_small_multitask_es_cs model
Model on translating legal text from Spanish to Cszech. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_es_cs model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Spanish to Cszech.
### How to use
Here is how to use this model to translate legal text from Spanish to Cszech in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_es_cs"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_es_cs", do_lower_case=False,
skip_special_tokens=True),
device=0
)
es_text = "La política pesquera supone que se tenga en cuenta un gran número de dimensiones – social, medioambiental, económica – lo que exige un enfoque integrado y equilibrado, incompatible con una visión que los sobrestima, en particular, mediante una definición a priori de cualquier jerarquía de prioridades."
pipeline([es_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_es_cs model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_es_cs | 47.673|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_en_sv | SEBIS | 2021-06-23T11:00:55Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation English Swedish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: English Swedish
tags:
- translation English Swedish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "whereas enlargement to Bulgaria and Romania should be effective in 2007,"
---
# legal_t5_small_multitask_en_sv model
Model on translating legal text from English to Swedish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_en_sv model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from English to Swedish.
### How to use
Here is how to use this model to translate legal text from English to Swedish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_en_sv"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_en_sv", do_lower_case=False,
skip_special_tokens=True),
device=0
)
en_text = "whereas enlargement to Bulgaria and Romania should be effective in 2007,"
pipeline([en_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_en_sv model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 9 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_en_sv | 47.968|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_en_fr | SEBIS | 2021-06-23T10:59:29Z | 6 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation English French model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: English French
tags:
- translation English French model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Article 2(b), sub-heading"
---
# legal_t5_small_multitask_en_fr model
Model on translating legal text from English to French. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_en_fr model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from English to French.
### How to use
Here is how to use this model to translate legal text from English to French in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_en_fr"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_en_fr", do_lower_case=False,
skip_special_tokens=True),
device=0
)
en_text = "Article 2(b), sub-heading"
pipeline([en_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_en_fr model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 9 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_en_fr | 38.063|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_de_sv | SEBIS | 2021-06-23T10:56:56Z | 6 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Deustch Swedish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Deustch Swedish
tags:
- translation Deustch Swedish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "SCHRIFTLICHE ANFRAGE P-1584/03"
---
# legal_t5_small_multitask_de_sv model
Model on translating legal text from Deustch to Swedish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_de_sv model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Deustch to Swedish.
### How to use
Here is how to use this model to translate legal text from Deustch to Swedish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_de_sv"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_de_sv", do_lower_case=False,
skip_special_tokens=True),
device=0
)
de_text = "SCHRIFTLICHE ANFRAGE P-1584/03"
pipeline([de_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_de_sv model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 8 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_de_sv | 35.945|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_de_es | SEBIS | 2021-06-23T10:54:59Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Deustch Spanish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Deustch Spanish
tags:
- translation Deustch Spanish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Kugelförmige, eiförmige oder ellipsenförmige Verpackungen dürfen keine Abmessungen aufweisen, die durch eine Einklemmung im Mund oder Rachen eine Blockierung der internen Atemwege verursachen können."
---
# legal_t5_small_multitask_de_es model
Model on translating legal text from Deustch to Spanish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_de_es model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Deustch to Spanish.
### How to use
Here is how to use this model to translate legal text from Deustch to Spanish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_de_es"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_de_es", do_lower_case=False,
skip_special_tokens=True),
device=0
)
de_text = "Kugelförmige, eiförmige oder ellipsenförmige Verpackungen dürfen keine Abmessungen aufweisen, die durch eine Einklemmung im Mund oder Rachen eine Blockierung der internen Atemwege verursachen können."
pipeline([de_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_de_es model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 8 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_de_es | 36.458|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_de_en | SEBIS | 2021-06-23T10:54:24Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Deustch English model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Deustch English
tags:
- translation Deustch English model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Der zuständige Ausschuss wacht darüber, dass alle Angaben, die die Ausübung des Mandats eines Mitglieds bzw. die Rangfolge der Stellvertreter beeinflussen können, dem Parlament unverzüglich von den Behörden der Mitgliedstaaten und der Union - unter Angabe deren Wirksamwerdens im Falle einer Benennung - übermittelt werden."
---
# legal_t5_small_multitask_de_en model
Model on translating legal text from Deustch to English. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_de_en model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Deustch to English.
### How to use
Here is how to use this model to translate legal text from Deustch to English in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_de_en"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_de_en", do_lower_case=False,
skip_special_tokens=True),
device=0
)
de_text = "Der zuständige Ausschuss wacht darüber, dass alle Angaben, die die Ausübung des Mandats eines Mitglieds bzw. die Rangfolge der Stellvertreter beeinflussen können, dem Parlament unverzüglich von den Behörden der Mitgliedstaaten und der Union - unter Angabe deren Wirksamwerdens im Falle einer Benennung - übermittelt werden."
pipeline([de_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_de_en model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 9 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_de_en | 42.437|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_cs_fr | SEBIS | 2021-06-23T10:52:26Z | 6 | 0 | transformers | [
"transformers",
"pytorch",
"t5",
"text2text-generation",
"translation Cszech French model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech French
tags:
- translation Cszech French model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Agentura USA pro ochranu životního prostředí ve své hodnotící studii v roce 2002 zjistila možnou systémovou toxicitu a karcinogenitu a údaje získané z krevních testů nasvědčují rozsáhlé expozici obyvatelstva."
---
# legal_t5_small_multitask_cs_fr model
Model on translating legal text from Cszech to French. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_cs_fr model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Cszech to French.
### How to use
Here is how to use this model to translate legal text from Cszech to French in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_cs_fr"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_cs_fr", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "Agentura USA pro ochranu životního prostředí ve své hodnotící studii v roce 2002 zjistila možnou systémovou toxicitu a karcinogenitu a údaje získané z krevních testů nasvědčují rozsáhlé expozici obyvatelstva."
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_cs_fr model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_cs_fr | 47.588|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_cs_es | SEBIS | 2021-06-23T10:51:58Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Cszech Spanish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech Spanish
tags:
- translation Cszech Spanish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Antonio Tajani (místopředseda Komise) ."
---
# legal_t5_small_multitask_cs_es model
Model on translating legal text from Cszech to Spanish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_cs_es model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Cszech to Spanish.
### How to use
Here is how to use this model to translate legal text from Cszech to Spanish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_cs_es"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_cs_es", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "Antonio Tajani (místopředseda Komise) ."
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_cs_es model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_cs_es | 48.559|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_multitask_cs_en | SEBIS | 2021-06-23T10:51:17Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Cszech English model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Cszech English
tags:
- translation Cszech English model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Komise musí vypracovat zprávu o hodnotících zprávách týkajících se uplatňování této směrnice v členských státech."
---
# legal_t5_small_multitask_cs_en model
Model on translating legal text from Cszech to English. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). The model is parallely trained on the three parallel corpus with 42 language pair
from jrc-acquis, europarl and dcep along with the unsupervised task where the model followed the task of prediction in a masked language model.
## Model description
No pretraining is involved in case of legal_t5_small_multitask_cs_en model, rather the unsupervised task is added with all the translation task
to realize the multitask learning scenario.
## Intended uses & limitations
The model could be used for translation of legal texts from Cszech to English.
### How to use
Here is how to use this model to translate legal text from Cszech to English in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_multitask_cs_en"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_multitask_cs_en", do_lower_case=False,
skip_special_tokens=True),
device=0
)
cs_text = "Komise musí vypracovat zprávu o hodnotících zprávách týkajících se uplatňování této směrnice v členských státech."
pipeline([cs_text], max_length=512)
```
## Training data
The legal_t5_small_multitask_cs_en model (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_multitask_cs_en | 37.136|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_cls_it | SEBIS | 2021-06-23T10:36:37Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"classification Italian model",
"dataset:jrc-acquis",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Italian
tags:
- classification Italian model
datasets:
- jrc-acquis
widget:
- text: "Regolamento (CE) n. 435/2005 della Commissione del 17 marzo 2005 relativo all'applicazione di un coefficiente di riduzione ai certificati di restituzione per le merci non comprese nell'allegato I del trattato come statuito all'articolo 8, paragrafo 5, del regolamento (CE) n. 1520/2000 LA COMMISSIONE DELLE COMUNITÀ EUROPEE, visto il trattato che istituisce la Comunità europea, visto il regolamento (CE) n. 3448/93 del Consiglio, del 6 dicembre 1993, sul regime di scambi per talune merci ottenute dalla trasformazione di prodotti agricoli [1], visto il regolamento (CE) n. 1520/2000 della Commissione, del 13 luglio 2000, che stabilisce, per taluni prodotti agricoli esportati sotto forma di merci non comprese nell'allegato I del trattato, le modalità comuni di applicazione relative al versamento delle restituzioni all'esportazione e i criteri per stabilirne l'importo [2], in particolare l'articolo 8, paragrafo 5, considerando quanto segue: (1) Dalle comunicazioni degli Stati membri di cui all'articolo 8, paragrafo 2, del regolamento (CE) n. 1520/2000 si evince che l'importo totale delle domande ricevute ammonta a 178002906 EUR, mentre l'importo disponibile per la tranche di titoli di restituzione di cui all'articolo 8, paragrafo 4, del regolamento (CE) n. 1520/2000 ammonta a 68116869 EUR. (2) Un coefficiente di riduzione è calcolato sulla base dell'articolo 8, paragrafi 3 e 4, del regolamento (CE) n. 1520/2000. Siffatto coefficiente dovrebbe pertanto essere applicato agli importi richiesti sotto forma di certificati di restituzione per il periodo dal 1o aprile 2005 come stabilito all'articolo 8, paragrafo 6, del regolamento (CE) n. 1520/2000, HA ADOTTATO IL PRESENTE REGOLAMENTO: Articolo 1 Gli importi delle domande di certificati di restituzione per il periodo dal 1o aprile 2005 sono soggetti a un coefficiente di riduzione pari a 0,618. Articolo 2 Il presente regolamento entra in vigore il 18 marzo 2005. Il presente regolamento è obbligatorio in tutti i suoi elementi e direttamente applicabile in ciascuno degli Stati membri. Fatto a Bruxelles, il 17 marzo 2005. Per la Commissione Günter Verheugen Vicepresidente [1] GU L 318 del 20.12.1993, pag. 18. Regolamento modificato da ultimo dal regolamento (CE) n. 2580/2000 (GU L 298 del 25.11.2000, pag. 5). [2] GU L 177 del 15.7.2000, pag. 1. Regolamento modificato da ultimo dal regolamento (CE) n. 886/2004 (GU L 168 del 1.5.2004, pag. 14). --------------------------------------------------"
---
# legal_t5_small_cls_it model
Model for classification of legal text written in Italian. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis.
## Model description
legal_t5_small_cls_it is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for classification of legal texts written in Italian.
### How to use
Here is how to use this model to classify legal text written in Italian in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_cls_it"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_cls_it", do_lower_case=False,
skip_special_tokens=True),
device=0
)
it_text = "Regolamento (CE) n. 435/2005 della Commissione del 17 marzo 2005 relativo all'applicazione di un coefficiente di riduzione ai certificati di restituzione per le merci non comprese nell'allegato I del trattato come statuito all'articolo 8, paragrafo 5, del regolamento (CE) n. 1520/2000 LA COMMISSIONE DELLE COMUNITÀ EUROPEE, visto il trattato che istituisce la Comunità europea, visto il regolamento (CE) n. 3448/93 del Consiglio, del 6 dicembre 1993, sul regime di scambi per talune merci ottenute dalla trasformazione di prodotti agricoli [1], visto il regolamento (CE) n. 1520/2000 della Commissione, del 13 luglio 2000, che stabilisce, per taluni prodotti agricoli esportati sotto forma di merci non comprese nell'allegato I del trattato, le modalità comuni di applicazione relative al versamento delle restituzioni all'esportazione e i criteri per stabilirne l'importo [2], in particolare l'articolo 8, paragrafo 5, considerando quanto segue: (1) Dalle comunicazioni degli Stati membri di cui all'articolo 8, paragrafo 2, del regolamento (CE) n. 1520/2000 si evince che l'importo totale delle domande ricevute ammonta a 178002906 EUR, mentre l'importo disponibile per la tranche di titoli di restituzione di cui all'articolo 8, paragrafo 4, del regolamento (CE) n. 1520/2000 ammonta a 68116869 EUR. (2) Un coefficiente di riduzione è calcolato sulla base dell'articolo 8, paragrafi 3 e 4, del regolamento (CE) n. 1520/2000. Siffatto coefficiente dovrebbe pertanto essere applicato agli importi richiesti sotto forma di certificati di restituzione per il periodo dal 1o aprile 2005 come stabilito all'articolo 8, paragrafo 6, del regolamento (CE) n. 1520/2000, HA ADOTTATO IL PRESENTE REGOLAMENTO: Articolo 1 Gli importi delle domande di certificati di restituzione per il periodo dal 1o aprile 2005 sono soggetti a un coefficiente di riduzione pari a 0,618. Articolo 2 Il presente regolamento entra in vigore il 18 marzo 2005. Il presente regolamento è obbligatorio in tutti i suoi elementi e direttamente applicabile in ciascuno degli Stati membri. Fatto a Bruxelles, il 17 marzo 2005. Per la Commissione Günter Verheugen Vicepresidente [1] GU L 318 del 20.12.1993, pag. 18. Regolamento modificato da ultimo dal regolamento (CE) n. 2580/2000 (GU L 298 del 25.11.2000, pag. 5). [2] GU L 177 del 15.7.2000, pag. 1. Regolamento modificato da ultimo dal regolamento (CE) n. 886/2004 (GU L 168 del 1.5.2004, pag. 14). --------------------------------------------------"
pipeline([it_text], max_length=512)
```
## Training data
The legal_t5_small_cls_it model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html) dataset consisting of 23 Thousand texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 64). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for classification test dataset, achieves the following results:
Test results :
| Model | F1 score |
|:-----:|:-----:|
| legal_t5_small_cls_it | 0.6296|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_cls_es | SEBIS | 2021-06-23T10:29:12Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"classification Spanish model",
"dataset:jrc-acquis",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Spanish
tags:
- classification Spanish model
datasets:
- jrc-acquis
widget:
- text: "Reglamento (CE) no 90/2001 de la Comisión de 17 de enero de 2001 que modifica el Reglamento (CE) n° 800/1999 por el que se establecen disposiciones comunes de aplicación del régimen de restituciones por exportación de productos agrícolas LA COMISIÓN DE LAS COMUNIDADES EUROPEAS, Visto el Tratado constitutivo de la Comunidad Europea, Visto el Reglamento (CEE) n° 1766/92 del Consejo, de 30 de junio de 1992, por el que se establece la organización común de mercados en el sector de los cereales(1), cuya última modificación la constituye el Reglamento (CE) n° 1666/2000(2), y, en particular, sus artículos 13 y 21, así como las disposiciones correspondientes de los demás Reglamentos por los que se establecen organizaciones comunes de mercados de productos agrícolas, Considerando lo siguiente: (1) En el caso de exportación de productos presentados a granel o en unidades no normalizadas, en los que es evidente que la masa neta exacta de los productos no puede conocerse hasta después de cargar el medio de transporte, el apartado 6 del artículo 5 del Reglamento (CE) n° 800/1999 de la Comisión(3), modificado por el Reglamento (CE) n° 1557/2000(4) establece la aplicación de una reducción de la restitución cuando la masa neta efectivamente cargada sea inferior a un determinado porcentaje de la masa neta estimada. No obstante, para la aplicación de esta disposición conviene tener en cuenta las limitaciones inherentes a los medios de transporte de navegación marítima o interior. En efecto, en el caso de los productos exportados a granel, puede ocurrir que las cantidades declaradas no se carguen en su totalidad debido, en particular, a la decisión del responsable del medio de transporte que puede ordenar la suspensión de la carga por razones técnicas o debido a un exceso de carga imputable a los demás exportadores. (2) Dado que determinados cortes de carne de porcino no se presentan en embalajes ni son, por naturaleza, homogéneos, conviene ampliar la categoría de unidades no normalizadas a este tipo de productos. (3) En lo que respecta a la noción de lugar de carga, en el comercio de exportación de productos agrícolas se presenta una multitud de situaciones comerciales y administrativas; por consiguiente, es difícil establecer una norma única y conviene autorizar a los Estados miembros para que determinen el lugar más apropiado para efectuar los controles físicos para los productos agrícolas exportados que se benefician de una restitución. A estos efectos, parece justificado determinar el lugar de carga, de forma diferente, en función de que los productos sean cargados en contenedores o, por el contrario, a granel, en sacos o en cajas y no se carguen posteriormente en contenedores. Asimismo, es conveniente que, cuando existan motivos debidamente justificados, se permita que las autoridades aduaneras acepten para los productos agrícolas que se beneficien, de una restitución declaraciones de exportación presentadas en una oficina de aduanas que no sea la del lugar donde vayan a cargarse los productos. (4) En el caso de los productos sujetos al régimen de mercancías de retorno, es oportuno prever la posibilidad de que la reintroducción se efectúe, bien por el Estado miembros del que sean originarios los productos, bien por el Estado miembro exportador de la primera exportación. (5) Conviene modificar el Reglamento (CE) n° 800/1999 en consecuencia. (6) Las medidas previstas en el presente Reglamento se ajustan al dictamen de todos los Comités de gestión interesados. HA ADOPTADO EL PRESENTE REGLAMENTO: Artículo 1 El Reglamento (CE) n° 800/1999 se modificará como sigue: 1) En el apartado 6 del articulo 5, el párrafo tercero se sustituirá por el texto siguiente: %quot%No se concederá ninguna restitución por la cantidad que sobrepase el 110 % de la masa neta estimada. Cuando la masa efectivamente cargada sea inferior al 90 % de la masa neta estimada, la restitución por la masa neta efectivamente cargada se reducirá un 10 % en relación con la diferencia entre la restitución correspondiente al 90 % de la masa neta estimada y la restitución correspondiente a la masa efectivamente cargada. No obstante, en los casos de exportación par vía marítima o por vía navegable interior, la restitución se pagará por la masa neta efectivamente cargada cuando el exportador pueda aportar la prueba, refrendada por el responsable del medio de transporte, de que el hecho de que no se cargara la totalidad de sus mercancías se debió a las limitaciones inherentes a ese tipo de transporte o a un exceso de carga imputable a uno o a varios de los demás exportadores. En caso de que el exportador haya utilizado el procedimiento de domiciliación previsto en el artículo 283 del Reglamento (CEE) n° 2454/93 serán aplicables las disposiciones del presente párrafo siempre que las autoridades aduaneras hayan autorizado la rectificación de los documentos contables en los que los productos exportados hayan sido inscritos.%quot%. 2) En el apartado 6 del artículo 5, el párrafo cuarto se sustituirá por el texto siguiente: %quot%Se considerarán productos en unidades no estandarizadas los animales vivos, las (medias) canales, los cuartos, partes delanteras, jamones, paletillas, pechos y lomos.%quot%. 3) El apartado 7 del articulo 5 se sustituirá por el texto siguiente: %quot%7. Cualquier persona que exporte productos por los cuales solicite la concesión de la restitución estará obligada a lo siguiente: a) presentar la declaración de exportación en la oficina de aduanas competente del lugar en que los productos vayan a cargarse en el transporte que vaya a efectuar la exportación; b) informar a dicha oficina de aduanas, coma mínimo 24 horas antes del comienzo de las operaciones de carga, e indicar la duración prevista de las operaciones de carga; las autoridades competentes podrán modificar el plazo de 24 horas. Se podrá considerar como lugar de carga en el transporte de los productos destinados a la exportación: - en el caso de los productos que se exporten cargados en contenedores, el lugar donde se carguen en éstos las mercancías, - en el caso de los productos que se exporten a granel, en sacos, cajones, cajas, botellas, etc. sin cargarse en contenedores, el lugar donde se cargue el medio de transporte por el que las mercancías vayan a salir del territorio aduanero de la Comunidad. La oficina de aduanas competente podrá autorizar las operaciones de carga una vez aceptada la declaración de exportación y antes de finalizar el plazo a que se refiere la letra b). La oficina de aduanas competente deberá estar en condiciones de realizar el control físico y de aplicar las medidas de identificación necesarias para el transporte hacia la oficina de salida del territorio aduanero de la Comunidad. Si por razones de organización administrativa o por otras razones debidamente justificadas, no pueden aplicarse las disposiciones del párrafo primero, la declaración de exportación, sólo podrá ser presentada en la oficina de aduanas competente del Estado miembro en cuestión, y, en el caso de un control físico de conformidad con el Reglamento (CEE) n° 386/90, el producto presentado deberá ser descargado completamente. No obstante, la descarga completa no será obligatoria cuando las autoridades competentes puedan garantizar la realización de un control físico exhaustivo.%quot%. 4) En el apartado 3 del artículo 25, el último párrafo se sustituirá por el texto siguiente: %quot%La presente disposición sólo se aplicará cuando el régimen de retorno haya sido utilizado en el Estado miembro donde se haya aceptado la declaración de exportación de la primera exportación o en el Estado miembro de origen, de conformidad con el artículo 15 de la Directiva 97/78/CE del Consejo(5), por la que se establecen los principios relativos a la organización de controles veterinarios de los productos que se introduzcan en la Comunidad procedentes de terceros países.%quot%. Artículo 2 El presente Reglamento entrará en vigor el séptimo día siguiente al de su publicación en el Diario Oficial de las Comunidades Europeas. A petición de los exportadores, las disposiciones del apartado 1 del articulo 1 se aplicarán a los expedientes de restituciones que aún no hayan sido cerrados en el momento de la entrada en vigor del presente Reglamento. El presente Reglamento será obligatorio en todos sus elementos y directamente aplicable en cada Estado miembro. Hecho en Bruselas, el 17 de enero de 2001. Por la Comisión Franz Fischler Miembro de la Comisión (1) DO L 181 de 1.7.1992, p. 21. (2) DO L 193 de 29.7.2000, p. 1. (3) DO L 102 de 17.4.1999, p. 11. (4) DO L 179 de 18.7.2000, p. 6. (5) DO L 24 de 30.1.1998, p. 9."
---
# legal_t5_small_cls_es model
Model for classification of legal text written in Spanish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis.
## Model description
legal_t5_small_cls_es is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for classification of legal texts written in Spanish.
### How to use
Here is how to use this model to classify legal text written in Spanish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_cls_es"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_cls_es", do_lower_case=False,
skip_special_tokens=True),
device=0
)
es_text = "Reglamento (CE) no 90/2001 de la Comisión de 17 de enero de 2001 que modifica el Reglamento (CE) n° 800/1999 por el que se establecen disposiciones comunes de aplicación del régimen de restituciones por exportación de productos agrícolas LA COMISIÓN DE LAS COMUNIDADES EUROPEAS, Visto el Tratado constitutivo de la Comunidad Europea, Visto el Reglamento (CEE) n° 1766/92 del Consejo, de 30 de junio de 1992, por el que se establece la organización común de mercados en el sector de los cereales(1), cuya última modificación la constituye el Reglamento (CE) n° 1666/2000(2), y, en particular, sus artículos 13 y 21, así como las disposiciones correspondientes de los demás Reglamentos por los que se establecen organizaciones comunes de mercados de productos agrícolas, Considerando lo siguiente: (1) En el caso de exportación de productos presentados a granel o en unidades no normalizadas, en los que es evidente que la masa neta exacta de los productos no puede conocerse hasta después de cargar el medio de transporte, el apartado 6 del artículo 5 del Reglamento (CE) n° 800/1999 de la Comisión(3), modificado por el Reglamento (CE) n° 1557/2000(4) establece la aplicación de una reducción de la restitución cuando la masa neta efectivamente cargada sea inferior a un determinado porcentaje de la masa neta estimada. No obstante, para la aplicación de esta disposición conviene tener en cuenta las limitaciones inherentes a los medios de transporte de navegación marítima o interior. En efecto, en el caso de los productos exportados a granel, puede ocurrir que las cantidades declaradas no se carguen en su totalidad debido, en particular, a la decisión del responsable del medio de transporte que puede ordenar la suspensión de la carga por razones técnicas o debido a un exceso de carga imputable a los demás exportadores. (2) Dado que determinados cortes de carne de porcino no se presentan en embalajes ni son, por naturaleza, homogéneos, conviene ampliar la categoría de unidades no normalizadas a este tipo de productos. (3) En lo que respecta a la noción de lugar de carga, en el comercio de exportación de productos agrícolas se presenta una multitud de situaciones comerciales y administrativas; por consiguiente, es difícil establecer una norma única y conviene autorizar a los Estados miembros para que determinen el lugar más apropiado para efectuar los controles físicos para los productos agrícolas exportados que se benefician de una restitución. A estos efectos, parece justificado determinar el lugar de carga, de forma diferente, en función de que los productos sean cargados en contenedores o, por el contrario, a granel, en sacos o en cajas y no se carguen posteriormente en contenedores. Asimismo, es conveniente que, cuando existan motivos debidamente justificados, se permita que las autoridades aduaneras acepten para los productos agrícolas que se beneficien, de una restitución declaraciones de exportación presentadas en una oficina de aduanas que no sea la del lugar donde vayan a cargarse los productos. (4) En el caso de los productos sujetos al régimen de mercancías de retorno, es oportuno prever la posibilidad de que la reintroducción se efectúe, bien por el Estado miembros del que sean originarios los productos, bien por el Estado miembro exportador de la primera exportación. (5) Conviene modificar el Reglamento (CE) n° 800/1999 en consecuencia. (6) Las medidas previstas en el presente Reglamento se ajustan al dictamen de todos los Comités de gestión interesados. HA ADOPTADO EL PRESENTE REGLAMENTO: Artículo 1 El Reglamento (CE) n° 800/1999 se modificará como sigue: 1) En el apartado 6 del articulo 5, el párrafo tercero se sustituirá por el texto siguiente: %quot%No se concederá ninguna restitución por la cantidad que sobrepase el 110 % de la masa neta estimada. Cuando la masa efectivamente cargada sea inferior al 90 % de la masa neta estimada, la restitución por la masa neta efectivamente cargada se reducirá un 10 % en relación con la diferencia entre la restitución correspondiente al 90 % de la masa neta estimada y la restitución correspondiente a la masa efectivamente cargada. No obstante, en los casos de exportación par vía marítima o por vía navegable interior, la restitución se pagará por la masa neta efectivamente cargada cuando el exportador pueda aportar la prueba, refrendada por el responsable del medio de transporte, de que el hecho de que no se cargara la totalidad de sus mercancías se debió a las limitaciones inherentes a ese tipo de transporte o a un exceso de carga imputable a uno o a varios de los demás exportadores. En caso de que el exportador haya utilizado el procedimiento de domiciliación previsto en el artículo 283 del Reglamento (CEE) n° 2454/93 serán aplicables las disposiciones del presente párrafo siempre que las autoridades aduaneras hayan autorizado la rectificación de los documentos contables en los que los productos exportados hayan sido inscritos.%quot%. 2) En el apartado 6 del artículo 5, el párrafo cuarto se sustituirá por el texto siguiente: %quot%Se considerarán productos en unidades no estandarizadas los animales vivos, las (medias) canales, los cuartos, partes delanteras, jamones, paletillas, pechos y lomos.%quot%. 3) El apartado 7 del articulo 5 se sustituirá por el texto siguiente: %quot%7. Cualquier persona que exporte productos por los cuales solicite la concesión de la restitución estará obligada a lo siguiente: a) presentar la declaración de exportación en la oficina de aduanas competente del lugar en que los productos vayan a cargarse en el transporte que vaya a efectuar la exportación; b) informar a dicha oficina de aduanas, coma mínimo 24 horas antes del comienzo de las operaciones de carga, e indicar la duración prevista de las operaciones de carga; las autoridades competentes podrán modificar el plazo de 24 horas. Se podrá considerar como lugar de carga en el transporte de los productos destinados a la exportación: - en el caso de los productos que se exporten cargados en contenedores, el lugar donde se carguen en éstos las mercancías, - en el caso de los productos que se exporten a granel, en sacos, cajones, cajas, botellas, etc. sin cargarse en contenedores, el lugar donde se cargue el medio de transporte por el que las mercancías vayan a salir del territorio aduanero de la Comunidad. La oficina de aduanas competente podrá autorizar las operaciones de carga una vez aceptada la declaración de exportación y antes de finalizar el plazo a que se refiere la letra b). La oficina de aduanas competente deberá estar en condiciones de realizar el control físico y de aplicar las medidas de identificación necesarias para el transporte hacia la oficina de salida del territorio aduanero de la Comunidad. Si por razones de organización administrativa o por otras razones debidamente justificadas, no pueden aplicarse las disposiciones del párrafo primero, la declaración de exportación, sólo podrá ser presentada en la oficina de aduanas competente del Estado miembro en cuestión, y, en el caso de un control físico de conformidad con el Reglamento (CEE) n° 386/90, el producto presentado deberá ser descargado completamente. No obstante, la descarga completa no será obligatoria cuando las autoridades competentes puedan garantizar la realización de un control físico exhaustivo.%quot%. 4) En el apartado 3 del artículo 25, el último párrafo se sustituirá por el texto siguiente: %quot%La presente disposición sólo se aplicará cuando el régimen de retorno haya sido utilizado en el Estado miembro donde se haya aceptado la declaración de exportación de la primera exportación o en el Estado miembro de origen, de conformidad con el artículo 15 de la Directiva 97/78/CE del Consejo(5), por la que se establecen los principios relativos a la organización de controles veterinarios de los productos que se introduzcan en la Comunidad procedentes de terceros países.%quot%. Artículo 2 El presente Reglamento entrará en vigor el séptimo día siguiente al de su publicación en el Diario Oficial de las Comunidades Europeas. A petición de los exportadores, las disposiciones del apartado 1 del articulo 1 se aplicarán a los expedientes de restituciones que aún no hayan sido cerrados en el momento de la entrada en vigor del presente Reglamento. El presente Reglamento será obligatorio en todos sus elementos y directamente aplicable en cada Estado miembro. Hecho en Bruselas, el 17 de enero de 2001. Por la Comisión Franz Fischler Miembro de la Comisión (1) DO L 181 de 1.7.1992, p. 21. (2) DO L 193 de 29.7.2000, p. 1. (3) DO L 102 de 17.4.1999, p. 11. (4) DO L 179 de 18.7.2000, p. 6. (5) DO L 24 de 30.1.1998, p. 9."
pipeline([es_text], max_length=512)
```
## Training data
The legal_t5_small_cls_es model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html) dataset consisting of 22 Thousand texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 64). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for classification test dataset, achieves the following results:
Test results :
| Model | F1 score |
|:-----:|:-----:|
| legal_t5_small_cls_es | 0.6318|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/code_trans_t5_small_transfer_learning_pretrain | SEBIS | 2021-06-23T10:26:44Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | feature-extraction | 2022-03-02T23:29:04Z | # CodeTrans transfer learning pre-trained model
Pretrained model on programming languages using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans).
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used transfer-learning pre-training on 7 unsupervised datasets in the software development domain.
The model was trained on a single TPU Pod V3-8 for half million steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
It could be used to fine-tune other tasks in the software development domain.
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_source_code_summarization_sql_transfer_learning_finetune | SEBIS | 2021-06-23T10:26:05Z | 14 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "select time ( col0 ) from tab0"
---
# CodeTrans model for source code summarization sql
Pretrained model on programming language sql using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized sql code functions: it works best with tokenized sql functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used transfer-learning pre-training on 7 unsupervised datasets in the software development domain. It is then fine-tuned on the source code summarization task for the sql code snippets.
## Intended uses & limitations
The model could be used to generate the description for the sql function or be fine-tuned on other sql code tasks. It can be used on unparsed and untokenized sql code. However, if the sql code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate sql function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_sql_transfer_learning_finetune"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_sql_transfer_learning_finetune", skip_special_tokens=True),
device=0
)
tokenized_code = "select time ( col0 ) from tab0"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/transfer%20learning%20fine-tuning/source%20code%20summarization/sql/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Transfer-learning Pretraining
The model was trained on a single TPU Pod V3-8 for 500,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Fine-tuning
This model was then fine-tuned on a single TPU Pod V2-8 for 1000 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing sql code.
## Evaluation results
For the source code summarization tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | SQL | C# |
| -------------------- | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 8.45 | 17.55 | 19.74 |
| CodeTrans-ST-Base | 9.12 | 15.00 | 18.65 |
| CodeTrans-TF-Small | 10.06 | 17.71 | 20.40 |
| CodeTrans-TF-Base | 10.94 | 17.66 | 21.12 |
| CodeTrans-TF-Large | 12.41 | 18.40 | 21.43 |
| CodeTrans-MT-Small | 13.11 | 19.15 | 22.39 |
| CodeTrans-MT-Base | **13.37** | 19.24 | 23.20 |
| CodeTrans-MT-Large | 13.24 | 19.40 | **23.57** |
| CodeTrans-MT-TF-Small | 12.10 | 18.25 | 22.03 |
| CodeTrans-MT-TF-Base | 10.64 | 16.91 | 21.40 |
| CodeTrans-MT-TF-Large | 12.14 | **19.98** | 21.10 |
| CODE-NN | -- | 18.40 | 20.50 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_source_code_summarization_sql_multitask_finetune | SEBIS | 2021-06-23T10:25:19Z | 20 | 3 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "select time ( col0 ) from tab0"
---
# CodeTrans model for source code summarization sql
Pretrained model on programming language sql using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized sql code functions: it works best with tokenized sql functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets. It is then fine-tuned on the source code summarization task for the sql code snippets.
## Intended uses & limitations
The model could be used to generate the description for the sql function or be fine-tuned on other sql code tasks. It can be used on unparsed and untokenized sql code. However, if the sql code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate sql function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_sql_multitask_finetune"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_sql_multitask_finetune", skip_special_tokens=True),
device=0
)
tokenized_code = "select time ( col0 ) from tab0"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/fine-tuning/source%20code%20summarization/sql/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 500,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Fine-tuning
This model was then fine-tuned on a single TPU Pod V2-8 for 1200 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing sql code.
## Evaluation results
For the source code summarization tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | SQL | C# |
| -------------------- | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 8.45 | 17.55 | 19.74 |
| CodeTrans-ST-Base | 9.12 | 15.00 | 18.65 |
| CodeTrans-TF-Small | 10.06 | 17.71 | 20.40 |
| CodeTrans-TF-Base | 10.94 | 17.66 | 21.12 |
| CodeTrans-TF-Large | 12.41 | 18.40 | 21.43 |
| CodeTrans-MT-Small | 13.11 | 19.15 | 22.39 |
| CodeTrans-MT-Base | **13.37** | 19.24 | 23.20 |
| CodeTrans-MT-Large | 13.24 | 19.40 | **23.57** |
| CodeTrans-MT-TF-Small | 12.10 | 18.25 | 22.03 |
| CodeTrans-MT-TF-Base | 10.64 | 16.91 | 21.40 |
| CodeTrans-MT-TF-Large | 12.14 | **19.98** | 21.10 |
| CODE-NN | -- | 18.40 | 20.50 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_source_code_summarization_sql_multitask | SEBIS | 2021-06-23T10:24:43Z | 9 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "select time ( col0 ) from tab0"
---
# CodeTrans model for source code summarization sql
Pretrained model on programming language sql using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized sql code functions: it works best with tokenized sql functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets.
## Intended uses & limitations
The model could be used to generate the description for the sql function or be fine-tuned on other sql code tasks. It can be used on unparsed and untokenized sql code. However, if the sql code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate sql function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_sql_multitask"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_sql_multitask", skip_special_tokens=True),
device=0
)
tokenized_code = "select time ( col0 ) from tab0"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/pre-training/source%20code%20summarization/sql/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 460,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
## Evaluation results
For the source code summarization tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | SQL | C# |
| -------------------- | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 8.45 | 17.55 | 19.74 |
| CodeTrans-ST-Base | 9.12 | 15.00 | 18.65 |
| CodeTrans-TF-Small | 10.06 | 17.71 | 20.40 |
| CodeTrans-TF-Base | 10.94 | 17.66 | 21.12 |
| CodeTrans-TF-Large | 12.41 | 18.40 | 21.43 |
| CodeTrans-MT-Small | 13.11 | 19.15 | 22.39 |
| CodeTrans-MT-Base | **13.37** | 19.24 | 23.20 |
| CodeTrans-MT-Large | 13.24 | 19.40 | **23.57** |
| CodeTrans-MT-TF-Small | 12.10 | 18.25 | 22.03 |
| CodeTrans-MT-TF-Base | 10.64 | 16.91 | 21.40 |
| CodeTrans-MT-TF-Large | 12.14 | **19.98** | 21.10 |
| CODE-NN | -- | 18.40 | 20.50 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_source_code_summarization_python_multitask_finetune | SEBIS | 2021-06-23T10:23:12Z | 65 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: '''with open ( CODE_STRING , CODE_STRING ) as in_file : buf = in_file . readlines ( ) with open ( CODE_STRING , CODE_STRING ) as out_file : for line in buf : if line == " ; Include this text " : line = line + " Include below " out_file . write ( line ) '''
---
# CodeTrans model for source code summarization python
Pretrained model on programming language python using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized python code functions: it works best with tokenized python functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets. It is then fine-tuned on the source code summarization task for the python code snippets.
## Intended uses & limitations
The model could be used to generate the description for the python function or be fine-tuned on other python code tasks. It can be used on unparsed and untokenized python code. However, if the python code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate python function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_python_multitask_finetune"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_python_multitask_finetune", skip_special_tokens=True),
device=0
)
tokenized_code = '''with open ( CODE_STRING , CODE_STRING ) as in_file : buf = in_file . readlines ( ) with open ( CODE_STRING , CODE_STRING ) as out_file : for line in buf : if line == " ; Include this text " : line = line + " Include below " out_file . write ( line ) '''
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/fine-tuning/source%20code%20summarization/python/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 500,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Fine-tuning
This model was then fine-tuned on a single TPU Pod V2-8 for 600 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing python code.
## Evaluation results
For the source code summarization tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | SQL | C# |
| -------------------- | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 8.45 | 17.55 | 19.74 |
| CodeTrans-ST-Base | 9.12 | 15.00 | 18.65 |
| CodeTrans-TF-Small | 10.06 | 17.71 | 20.40 |
| CodeTrans-TF-Base | 10.94 | 17.66 | 21.12 |
| CodeTrans-TF-Large | 12.41 | 18.40 | 21.43 |
| CodeTrans-MT-Small | 13.11 | 19.15 | 22.39 |
| CodeTrans-MT-Base | **13.37** | 19.24 | 23.20 |
| CodeTrans-MT-Large | 13.24 | 19.40 | **23.57** |
| CodeTrans-MT-TF-Small | 12.10 | 18.25 | 22.03 |
| CodeTrans-MT-TF-Base | 10.64 | 16.91 | 21.40 |
| CodeTrans-MT-TF-Large | 12.14 | **19.98** | 21.10 |
| CODE-NN | -- | 18.40 | 20.50 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_source_code_summarization_python_multitask | SEBIS | 2021-06-23T10:22:36Z | 13 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: '''with open ( CODE_STRING , CODE_STRING ) as in_file : buf = in_file . readlines ( ) with open ( CODE_STRING , CODE_STRING ) as out_file : for line in buf : if line == " ; Include this text " : line = line + " Include below " out_file . write ( line ) '''
---
# CodeTrans model for source code summarization python
Pretrained model on programming language python using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized python code functions: it works best with tokenized python functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets.
## Intended uses & limitations
The model could be used to generate the description for the python function or be fine-tuned on other python code tasks. It can be used on unparsed and untokenized python code. However, if the python code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate python function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_python_multitask"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_python_multitask", skip_special_tokens=True),
device=0
)
tokenized_code = '''with open ( CODE_STRING , CODE_STRING ) as in_file : buf = in_file . readlines ( ) with open ( CODE_STRING , CODE_STRING ) as out_file : for line in buf : if line == " ; Include this text " : line = line + " Include below " out_file . write ( line ) '''
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/pre-training/source%20code%20summarization/python/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 300,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
## Evaluation results
For the source code summarization tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | SQL | C# |
| -------------------- | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 8.45 | 17.55 | 19.74 |
| CodeTrans-ST-Base | 9.12 | 15.00 | 18.65 |
| CodeTrans-TF-Small | 10.06 | 17.71 | 20.40 |
| CodeTrans-TF-Base | 10.94 | 17.66 | 21.12 |
| CodeTrans-TF-Large | 12.41 | 18.40 | 21.43 |
| CodeTrans-MT-Small | 13.11 | 19.15 | 22.39 |
| CodeTrans-MT-Base | **13.37** | 19.24 | 23.20 |
| CodeTrans-MT-Large | 13.24 | 19.40 | **23.57** |
| CodeTrans-MT-TF-Small | 12.10 | 18.25 | 22.03 |
| CodeTrans-MT-TF-Base | 10.64 | 16.91 | 21.40 |
| CodeTrans-MT-TF-Large | 12.14 | **19.98** | 21.10 |
| CODE-NN | -- | 18.40 | 20.50 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_source_code_summarization_csharp_transfer_learning_finetune | SEBIS | 2021-06-23T10:21:27Z | 11 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "public static DateTime ParseUnixDateTime ( double unixTime ) { var dt = new DateTime ( CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , System . DateTimeKind . Utc ) ; dt = dt . AddSeconds ( unixTimeStamp ) . ToLocalTime ( ) ; return dt ; }"
---
# CodeTrans model for source code summarization csharp
Pretrained model on programming language csharp using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized csharp code functions: it works best with tokenized csharp functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used transfer-learning pre-training on 7 unsupervised datasets in the software development domain. It is then fine-tuned on the source code summarization task for the csharp code snippets.
## Intended uses & limitations
The model could be used to generate the description for the csharp function or be fine-tuned on other csharp code tasks. It can be used on unparsed and untokenized csharp code. However, if the csharp code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate csharp function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_csharp_transfer_learning_finetune"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_csharp_transfer_learning_finetune", skip_special_tokens=True),
device=0
)
tokenized_code = "public static DateTime ParseUnixDateTime ( double unixTime ) { var dt = new DateTime ( CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , System . DateTimeKind . Utc ) ; dt = dt . AddSeconds ( unixTimeStamp ) . ToLocalTime ( ) ; return dt ; }"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/transfer%20learning%20fine-tuning/source%20code%20summarization/csharp/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Transfer-learning Pretraining
The model was trained on a single TPU Pod V3-8 for 500,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Fine-tuning
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 csharp code.
## Evaluation results
For the source code summarization tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | SQL | C# |
| -------------------- | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 8.45 | 17.55 | 19.74 |
| CodeTrans-ST-Base | 9.12 | 15.00 | 18.65 |
| CodeTrans-TF-Small | 10.06 | 17.71 | 20.40 |
| CodeTrans-TF-Base | 10.94 | 17.66 | 21.12 |
| CodeTrans-TF-Large | 12.41 | 18.40 | 21.43 |
| CodeTrans-MT-Small | 13.11 | 19.15 | 22.39 |
| CodeTrans-MT-Base | **13.37** | 19.24 | 23.20 |
| CodeTrans-MT-Large | 13.24 | 19.40 | **23.57** |
| CodeTrans-MT-TF-Small | 12.10 | 18.25 | 22.03 |
| CodeTrans-MT-TF-Base | 10.64 | 16.91 | 21.40 |
| CodeTrans-MT-TF-Large | 12.14 | **19.98** | 21.10 |
| CODE-NN | -- | 18.40 | 20.50 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_source_code_summarization_csharp_multitask_finetune | SEBIS | 2021-06-23T10:20:50Z | 11 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "public static DateTime ParseUnixDateTime ( double unixTime ) { var dt = new DateTime ( CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , System . DateTimeKind . Utc ) ; dt = dt . AddSeconds ( unixTimeStamp ) . ToLocalTime ( ) ; return dt ; }"
---
# CodeTrans model for source code summarization csharp
Pretrained model on programming language csharp using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized csharp code functions: it works best with tokenized csharp functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets. It is then fine-tuned on the source code summarization task for the csharp code snippets.
## Intended uses & limitations
The model could be used to generate the description for the csharp function or be fine-tuned on other csharp code tasks. It can be used on unparsed and untokenized csharp code. However, if the csharp code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate csharp function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_csharp_multitask_finetune"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_csharp_multitask_finetune", skip_special_tokens=True),
device=0
)
tokenized_code = "public static DateTime ParseUnixDateTime ( double unixTime ) { var dt = new DateTime ( CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , System . DateTimeKind . Utc ) ; dt = dt . AddSeconds ( unixTimeStamp ) . ToLocalTime ( ) ; return dt ; }"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/fine-tuning/source%20code%20summarization/csharp/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 500,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Fine-tuning
This model was then fine-tuned on a single TPU Pod V2-8 for 1200 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing csharp code.
## Evaluation results
For the source code summarization tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | SQL | C# |
| -------------------- | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 8.45 | 17.55 | 19.74 |
| CodeTrans-ST-Base | 9.12 | 15.00 | 18.65 |
| CodeTrans-TF-Small | 10.06 | 17.71 | 20.40 |
| CodeTrans-TF-Base | 10.94 | 17.66 | 21.12 |
| CodeTrans-TF-Large | 12.41 | 18.40 | 21.43 |
| CodeTrans-MT-Small | 13.11 | 19.15 | 22.39 |
| CodeTrans-MT-Base | **13.37** | 19.24 | 23.20 |
| CodeTrans-MT-Large | 13.24 | 19.40 | **23.57** |
| CodeTrans-MT-TF-Small | 12.10 | 18.25 | 22.03 |
| CodeTrans-MT-TF-Base | 10.64 | 16.91 | 21.40 |
| CodeTrans-MT-TF-Large | 12.14 | **19.98** | 21.10 |
| CODE-NN | -- | 18.40 | 20.50 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_source_code_summarization_csharp_multitask | SEBIS | 2021-06-23T10:20:16Z | 14 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "public static DateTime ParseUnixDateTime ( double unixTime ) { var dt = new DateTime ( CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , System . DateTimeKind . Utc ) ; dt = dt . AddSeconds ( unixTimeStamp ) . ToLocalTime ( ) ; return dt ; }"
---
# CodeTrans model for source code summarization csharp
Pretrained model on programming language csharp using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized csharp code functions: it works best with tokenized csharp functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets.
## Intended uses & limitations
The model could be used to generate the description for the csharp function or be fine-tuned on other csharp code tasks. It can be used on unparsed and untokenized csharp code. However, if the csharp code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate csharp function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_csharp_multitask"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_source_code_summarization_csharp_multitask", skip_special_tokens=True),
device=0
)
tokenized_code = "public static DateTime ParseUnixDateTime ( double unixTime ) { var dt = new DateTime ( CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , CODE_INTEGER , System . DateTimeKind . Utc ) ; dt = dt . AddSeconds ( unixTimeStamp ) . ToLocalTime ( ) ; return dt ; }"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/pre-training/source%20code%20summarization/csharp/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 300,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
## Evaluation results
For the source code summarization tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | SQL | C# |
| -------------------- | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 8.45 | 17.55 | 19.74 |
| CodeTrans-ST-Base | 9.12 | 15.00 | 18.65 |
| CodeTrans-TF-Small | 10.06 | 17.71 | 20.40 |
| CodeTrans-TF-Base | 10.94 | 17.66 | 21.12 |
| CodeTrans-TF-Large | 12.41 | 18.40 | 21.43 |
| CodeTrans-MT-Small | 13.11 | 19.15 | 22.39 |
| CodeTrans-MT-Base | **13.37** | 19.24 | 23.20 |
| CodeTrans-MT-Large | 13.24 | 19.40 | **23.57** |
| CodeTrans-MT-TF-Small | 12.10 | 18.25 | 22.03 |
| CodeTrans-MT-TF-Base | 10.64 | 16.91 | 21.40 |
| CodeTrans-MT-TF-Large | 12.14 | **19.98** | 21.10 |
| CODE-NN | -- | 18.40 | 20.50 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_program_synthese_multitask_finetune | SEBIS | 2021-06-23T10:17:40Z | 13 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "you are given an array of numbers a and a number b , compute the difference of elements in a and b"
---
# CodeTrans model for program synthesis
Pretrained model on programming language lisp inspired DSL using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans).
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets. It is then fine-tuned on the program synthesis task for the lisp inspired DSL code.
## Intended uses & limitations
The model could be used to generate lisp inspired DSL code given the human language description tasks.
### How to use
Here is how to use this model to generate lisp inspired DSL code using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_program_synthese_multitask_finetune"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_program_synthese_multitask_finetune", skip_special_tokens=True),
device=0
)
tokenized_code = "you are given an array of numbers a and a number b , compute the difference of elements in a and b"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/fine-tuning/program%20synthesis/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 500,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Fine-tuning
This model was then fine-tuned on a single TPU Pod V2-8 for 16,000 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing lisp inspired DSL data.
## Evaluation results
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | LISP |
| -------------------- | :------------: |
| CodeTrans-ST-Small | 89.43 |
| CodeTrans-ST-Base | 89.65 |
| CodeTrans-TF-Small | 90.30 |
| CodeTrans-TF-Base | 90.24 |
| CodeTrans-TF-Large | 90.21 |
| CodeTrans-MT-Small | 82.88 |
| CodeTrans-MT-Base | 86.99 |
| CodeTrans-MT-Large | 90.27 |
| CodeTrans-MT-TF-Small | **90.31** |
| CodeTrans-MT-TF-Base | 90.30 |
| CodeTrans-MT-TF-Large | 90.17 |
| State of the art | 85.80 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_commit_generation_multitask_finetune | SEBIS | 2021-06-23T10:15:17Z | 11 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "new file mode 100644 index 000000000 . . 892fda21b Binary files / dev / null and b / src / plugins / gateway / lib / joscar . jar differ"
---
# CodeTrans model for git commit message generation
Pretrained model on git commit using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized git commit: it works best with tokenized git commit.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets. It is then fine-tuned on the git commit message generation task for the java commit changes.
## Intended uses & limitations
The model could be used to generate the git commit message for the git commit changes or be fine-tuned on other relevant tasks. It can be used on unparsed and untokenized commit changes. However, if the change is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate git commit message using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_commit_generation_multitask_finetune"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_commit_generation_multitask_finetune", skip_special_tokens=True),
device=0
)
tokenized_code = "new file mode 100644 index 000000000 . . 892fda21b Binary files / dev / null and b / src / plugins / gateway / lib / joscar . jar differ"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/fine-tuning/commit%20generation/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 500,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Fine-tuning
This model was then fine-tuned on a single TPU Pod V2-8 for 8,000 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing commit changes.
## Evaluation results
For the git commit message generation task, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Java |
| -------------------- | :------------: |
| CodeTrans-ST-Small | 39.61 |
| CodeTrans-ST-Base | 38.67 |
| CodeTrans-TF-Small | 44.22 |
| CodeTrans-TF-Base | 44.17 |
| CodeTrans-TF-Large | **44.41** |
| CodeTrans-MT-Small | 36.17 |
| CodeTrans-MT-Base | 39.25 |
| CodeTrans-MT-Large | 41.18 |
| CodeTrans-MT-TF-Small | 43.96 |
| CodeTrans-MT-TF-Base | 44.19 |
| CodeTrans-MT-TF-Large | 44.34 |
| State of the art | 32.81 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_commit_generation_multitask | SEBIS | 2021-06-23T10:14:38Z | 10 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "new file mode 100644 index 000000000 . . 892fda21b Binary files / dev / null and b / src / plugins / gateway / lib / joscar . jar differ"
---
# CodeTrans model for git commit message generation
Pretrained model on git commit using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized git commit: it works best with tokenized git commit.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets.
## Intended uses & limitations
The model could be used to generate the git commit message for the git commit changes or be fine-tuned on other relevant tasks. It can be used on unparsed and untokenized commit changes. However, if the change is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate git commit message using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_commit_generation_multitask"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_commit_generation_multitask", skip_special_tokens=True),
device=0
)
tokenized_code = "new file mode 100644 index 000000000 . . 892fda21b Binary files / dev / null and b / src / plugins / gateway / lib / joscar . jar differ"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/pre-training/commit%20generation/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 360,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
## Evaluation results
For the git commit message generation task, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Java |
| -------------------- | :------------: |
| CodeTrans-ST-Small | 39.61 |
| CodeTrans-ST-Base | 38.67 |
| CodeTrans-TF-Small | 44.22 |
| CodeTrans-TF-Base | 44.17 |
| CodeTrans-TF-Large | **44.41** |
| CodeTrans-MT-Small | 36.17 |
| CodeTrans-MT-Base | 39.25 |
| CodeTrans-MT-Large | 41.18 |
| CodeTrans-MT-TF-Small | 43.96 |
| CodeTrans-MT-TF-Base | 44.19 |
| CodeTrans-MT-TF-Large | 44.34 |
| State of the art | 32.81 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_code_documentation_generation_ruby_transfer_learning_finetune | SEBIS | 2021-06-23T10:13:18Z | 10 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "def add ( severity , progname , & block ) return true if io . nil? || severity < level message = format_message ( severity , progname , yield ) MUTEX . synchronize { io . write ( message ) } true end"
---
# CodeTrans model for code documentation generation ruby
Pretrained model on programming language ruby using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized ruby code functions: it works best with tokenized ruby functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used transfer-learning pre-training on 7 unsupervised datasets in the software development domain. It is then fine-tuned on the code documentation generation task for the ruby function/method.
## Intended uses & limitations
The model could be used to generate the description for the ruby function or be fine-tuned on other ruby code tasks. It can be used on unparsed and untokenized ruby code. However, if the ruby code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate ruby function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_ruby_transfer_learning_finetune"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_ruby_transfer_learning_finetune", skip_special_tokens=True),
device=0
)
tokenized_code = "def add ( severity , progname , & block ) return true if io . nil? || severity < level message = format_message ( severity , progname , yield ) MUTEX . synchronize { io . write ( message ) } true end"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/transfer%20learning%20fine-tuning/function%20documentation%20generation/ruby/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Transfer-learning Pretraining
The model was trained on a single TPU Pod V3-8 for half million steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Fine-tuning
This model was then fine-tuned on a single TPU Pod V2-8 for 5000 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing ruby code.
## Evaluation results
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
| -------------------- | :------------: | :------------: | :------------: | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 17.31 | 16.65 | 16.89 | 23.05 | 9.19 | 13.7 |
| CodeTrans-ST-Base | 16.86 | 17.17 | 17.16 | 22.98 | 8.23 | 13.17 |
| CodeTrans-TF-Small | 19.93 | 19.48 | 18.88 | 25.35 | 13.15 | 17.23 |
| CodeTrans-TF-Base | 20.26 | 20.19 | 19.50 | 25.84 | 14.07 | 18.25 |
| CodeTrans-TF-Large | 20.35 | 20.06 | **19.54** | 26.18 | 14.94 | **18.98** |
| CodeTrans-MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
| CodeTrans-MT-Base | **20.39** | 21.22 | 19.43 | **26.23** | **15.26** | 16.11 |
| CodeTrans-MT-Large | 20.18 | **21.87** | 19.38 | 26.08 | 15.00 | 16.23 |
| CodeTrans-MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
| CodeTrans-MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | 18.62 |
| CodeTrans-MT-TF-Large | 18.94 | 21.42 | 18.77 | 26.20 | 14.19 | 18.83 |
| State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_code_documentation_generation_ruby_multitask_finetune | SEBIS | 2021-06-23T10:12:41Z | 10 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "def add ( severity , progname , & block ) return true if io . nil? || severity < level message = format_message ( severity , progname , yield ) MUTEX . synchronize { io . write ( message ) } true end"
---
# CodeTrans model for code documentation generation ruby
Pretrained model on programming language ruby using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized ruby code functions: it works best with tokenized ruby functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets. It is then fine-tuned on the code documentation generation task for the ruby function/method.
## Intended uses & limitations
The model could be used to generate the description for the ruby function or be fine-tuned on other ruby code tasks. It can be used on unparsed and untokenized ruby code. However, if the ruby code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate ruby function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_ruby_multitask_finetune"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_ruby_multitask_finetune", skip_special_tokens=True),
device=0
)
tokenized_code = "def add ( severity , progname , & block ) return true if io . nil? || severity < level message = format_message ( severity , progname , yield ) MUTEX . synchronize { io . write ( message ) } true end"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/fine-tuning/function%20documentation%20generation/ruby/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for half million steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Fine-tuning
This model was then fine-tuned on a single TPU Pod V2-8 for 2,000 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing ruby code.
## Evaluation results
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
| -------------------- | :------------: | :------------: | :------------: | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 17.31 | 16.65 | 16.89 | 23.05 | 9.19 | 13.7 |
| CodeTrans-ST-Base | 16.86 | 17.17 | 17.16 | 22.98 | 8.23 | 13.17 |
| CodeTrans-TF-Small | 19.93 | 19.48 | 18.88 | 25.35 | 13.15 | 17.23 |
| CodeTrans-TF-Base | 20.26 | 20.19 | 19.50 | 25.84 | 14.07 | 18.25 |
| CodeTrans-TF-Large | 20.35 | 20.06 | **19.54** | 26.18 | 14.94 | **18.98** |
| CodeTrans-MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
| CodeTrans-MT-Base | **20.39** | 21.22 | 19.43 | **26.23** | **15.26** | 16.11 |
| CodeTrans-MT-Large | 20.18 | **21.87** | 19.38 | 26.08 | 15.00 | 16.23 |
| CodeTrans-MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
| CodeTrans-MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | 18.62 |
| CodeTrans-MT-TF-Large | 18.94 | 21.42 | 18.77 | 26.20 | 14.19 | 18.83 |
| State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_code_documentation_generation_ruby_multitask | SEBIS | 2021-06-23T10:12:08Z | 13 | 1 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "def add ( severity , progname , & block ) return true if io . nil? || severity < level message = format_message ( severity , progname , yield ) MUTEX . synchronize { io . write ( message ) } true end"
---
# CodeTrans model for code documentation generation ruby
Pretrained model on programming language ruby using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized ruby code functions: it works best with tokenized ruby functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets.
## Intended uses & limitations
The model could be used to generate the description for the ruby function or be fine-tuned on other ruby code tasks. It can be used on unparsed and untokenized ruby code. However, if the ruby code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate ruby function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_ruby_multitask"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_ruby_multitask", skip_special_tokens=True),
device=0
)
tokenized_code = "def add ( severity , progname , & block ) return true if io . nil? || severity < level message = format_message ( severity , progname , yield ) MUTEX . synchronize { io . write ( message ) } true end"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/pre-training/function%20documentation%20generation/ruby/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 420,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
## Evaluation results
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
| -------------------- | :------------: | :------------: | :------------: | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 17.31 | 16.65 | 16.89 | 23.05 | 9.19 | 13.7 |
| CodeTrans-ST-Base | 16.86 | 17.17 | 17.16 | 22.98 | 8.23 | 13.17 |
| CodeTrans-TF-Small | 19.93 | 19.48 | 18.88 | 25.35 | 13.15 | 17.23 |
| CodeTrans-TF-Base | 20.26 | 20.19 | 19.50 | 25.84 | 14.07 | 18.25 |
| CodeTrans-TF-Large | 20.35 | 20.06 | **19.54** | 26.18 | 14.94 | **18.98** |
| CodeTrans-MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
| CodeTrans-MT-Base | **20.39** | 21.22 | 19.43 | **26.23** | **15.26** | 16.11 |
| CodeTrans-MT-Large | 20.18 | **21.87** | 19.38 | 26.08 | 15.00 | 16.23 |
| CodeTrans-MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
| CodeTrans-MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | 18.62 |
| CodeTrans-MT-TF-Large | 18.94 | 21.42 | 18.77 | 26.20 | 14.19 | 18.83 |
| State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_code_documentation_generation_ruby | SEBIS | 2021-06-23T10:11:41Z | 12 | 0 | transformers | [
"transformers",
"pytorch",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "def add ( severity , progname , & block ) return true if io . nil? || severity < level message = format_message ( severity , progname , yield ) MUTEX . synchronize { io . write ( message ) } true end"
---
# CodeTrans model for code documentation generation ruby
Pretrained model on programming language ruby using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized ruby code functions: it works best with tokenized ruby functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used single-task training on CodeSearchNet Corpus ruby dataset.
## Intended uses & limitations
The model could be used to generate the description for the ruby function or be fine-tuned on other ruby code tasks. It can be used on unparsed and untokenized ruby code. However, if the ruby code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate ruby function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_ruby"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_ruby", skip_special_tokens=True),
device=0
)
tokenized_code = "def add ( severity , progname , & block ) return true if io . nil? || severity < level message = format_message ( severity , progname , yield ) MUTEX . synchronize { io . write ( message ) } true end"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/single%20task/function%20documentation%20generation/ruby/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Evaluation results
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
| -------------------- | :------------: | :------------: | :------------: | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 17.31 | 16.65 | 16.89 | 23.05 | 9.19 | 13.7 |
| CodeTrans-ST-Base | 16.86 | 17.17 | 17.16 | 22.98 | 8.23 | 13.17 |
| CodeTrans-TF-Small | 19.93 | 19.48 | 18.88 | 25.35 | 13.15 | 17.23 |
| CodeTrans-TF-Base | 20.26 | 20.19 | 19.50 | 25.84 | 14.07 | 18.25 |
| CodeTrans-TF-Large | 20.35 | 20.06 | **19.54** | 26.18 | 14.94 | **18.98** |
| CodeTrans-MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
| CodeTrans-MT-Base | **20.39** | 21.22 | 19.43 | **26.23** | **15.26** | 16.11 |
| CodeTrans-MT-Large | 20.18 | **21.87** | 19.38 | 26.08 | 15.00 | 16.23 |
| CodeTrans-MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
| CodeTrans-MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | 18.62 |
| CodeTrans-MT-TF-Large | 18.94 | 21.42 | 18.77 | 26.20 | 14.19 | 18.83 |
| State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/legal_t5_small_trans_sv_fr_small_finetuned | SEBIS | 2021-06-23T10:11:10Z | 7 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Swedish French model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Swedish French
tags:
- translation Swedish French model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Samreglering bör följa samma principer som de formella bestämmelserna, vilket betyder att den bör vara objektiv, välgrundad, proportionell och icke-diskriminerande, och bör möjliggöra insyn."
---
# legal_t5_small_trans_sv_fr_small_finetuned model
Model on translating legal text from Swedish to French. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is first pretrained all the translation data over some unsupervised task. Then the model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_sv_fr_small_finetuned is initially pretrained on unsupervised task with the all of the data of the training set. The unsupervised task was "masked language modelling". legal_t5_small_trans_sv_fr_small_finetuned is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Swedish to French.
### How to use
Here is how to use this model to translate legal text from Swedish to French in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_sv_fr_small_finetuned"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_sv_fr", do_lower_case=False,
skip_special_tokens=True),
device=0
)
sv_text = "Samreglering bör följa samma principer som de formella bestämmelserna, vilket betyder att den bör vara objektiv, välgrundad, proportionell och icke-diskriminerande, och bör möjliggöra insyn."
pipeline([sv_text], max_length=512)
```
## Training data
The legal_t5_small_trans_sv_fr_small_finetuned (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 8 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
The pre-training data was the combined data from all the 42 language pairs. The task for the model was to predict the portions of a sentence which were masked randomly.
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_sv_fr_small_finetuned | 47.508|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_sv_fr | SEBIS | 2021-06-23T10:10:34Z | 7 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Swedish French model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Swedish French
tags:
- translation Swedish French model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Kunden måste ha rätt att avsäga sig information i skriftlig form."
---
# legal_t5_small_trans_sv_fr model
Model on translating legal text from Swedish to French. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_sv_fr is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Swedish to French.
### How to use
Here is how to use this model to translate legal text from Swedish to French in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_sv_fr"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_sv_fr", do_lower_case=False,
skip_special_tokens=True),
device=0
)
sv_text = "Kunden måste ha rätt att avsäga sig information i skriftlig form."
pipeline([sv_text], max_length=512)
```
## Training data
The legal_t5_small_trans_sv_fr model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_sv_fr | 47.623|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_sv_es_small_finetuned | SEBIS | 2021-06-23T10:09:55Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Swedish Spanish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Swedish Spanish
tags:
- translation Swedish Spanish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "– med beaktande av kommissionen vitbok om idrott ( KOM(2007)0391 ),"
---
# legal_t5_small_trans_sv_es_small_finetuned model
Model on translating legal text from Swedish to Spanish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is first pretrained all the translation data over some unsupervised task. Then the model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_sv_es_small_finetuned is initially pretrained on unsupervised task with the all of the data of the training set. The unsupervised task was "masked language modelling". legal_t5_small_trans_sv_es_small_finetuned is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Swedish to Spanish.
### How to use
Here is how to use this model to translate legal text from Swedish to Spanish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_sv_es_small_finetuned"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_sv_es", do_lower_case=False,
skip_special_tokens=True),
device=0
)
sv_text = "– med beaktande av kommissionen vitbok om idrott ( KOM(2007)0391 ),"
pipeline([sv_text], max_length=512)
```
## Training data
The legal_t5_small_trans_sv_es_small_finetuned (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 8 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
The pre-training data was the combined data from all the 42 language pairs. The task for the model was to predict the portions of a sentence which were masked randomly.
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_sv_es_small_finetuned | 47.411|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/code_trans_t5_small_code_documentation_generation_python | SEBIS | 2021-06-23T10:09:35Z | 17 | 1 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "def e ( message , exit_code = None ) : print_log ( message , YELLOW , BOLD ) if exit_code is not None : sys . exit ( exit_code )"
---
# CodeTrans model for code documentation generation python
Pretrained model on programming language python using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized python code functions: it works best with tokenized python functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used single-task training on CodeSearchNet Corpus python dataset.
## Intended uses & limitations
The model could be used to generate the description for the python function or be fine-tuned on other python code tasks. It can be used on unparsed and untokenized python code. However, if the python code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate python function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_python"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_python", skip_special_tokens=True),
device=0
)
tokenized_code = "def e ( message , exit_code = None ) : print_log ( message , YELLOW , BOLD ) if exit_code is not None : sys . exit ( exit_code )"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/single%20task/function%20documentation%20generation/python/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Evaluation results
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
| -------------------- | :------------: | :------------: | :------------: | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 17.31 | 16.65 | 16.89 | 23.05 | 9.19 | 13.7 |
| CodeTrans-ST-Base | 16.86 | 17.17 | 17.16 | 22.98 | 8.23 | 13.17 |
| CodeTrans-TF-Small | 19.93 | 19.48 | 18.88 | 25.35 | 13.15 | 17.23 |
| CodeTrans-TF-Base | 20.26 | 20.19 | 19.50 | 25.84 | 14.07 | 18.25 |
| CodeTrans-TF-Large | 20.35 | 20.06 | **19.54** | 26.18 | 14.94 | **18.98** |
| CodeTrans-MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
| CodeTrans-MT-Base | **20.39** | 21.22 | 19.43 | **26.23** | **15.26** | 16.11 |
| CodeTrans-MT-Large | 20.18 | **21.87** | 19.38 | 26.08 | 15.00 | 16.23 |
| CodeTrans-MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
| CodeTrans-MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | 18.62 |
| CodeTrans-MT-TF-Large | 18.94 | 21.42 | 18.77 | 26.20 | 14.19 | 18.83 |
| State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/legal_t5_small_trans_sv_es | SEBIS | 2021-06-23T10:09:20Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Swedish Spanish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Swedish Spanish
tags:
- translation Swedish Spanish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Monika Flašíková Beňová (S&D)"
---
# legal_t5_small_trans_sv_es model
Model on translating legal text from Swedish to Spanish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_sv_es is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Swedish to Spanish.
### How to use
Here is how to use this model to translate legal text from Swedish to Spanish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_sv_es"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_sv_es", do_lower_case=False,
skip_special_tokens=True),
device=0
)
sv_text = "Monika Flašíková Beňová (S&D)"
pipeline([sv_text], max_length=512)
```
## Training data
The legal_t5_small_trans_sv_es model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_sv_es | 47.407|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_sv_en | SEBIS | 2021-06-23T10:08:13Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Swedish English model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Swedish English
tags:
- translation Swedish English model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Om rättsliga förfaranden inleds rörande omständigheter som ombudsmannen utreder skall han avsluta ärendet."
---
# legal_t5_small_trans_sv_en model
Model on translating legal text from Swedish to English. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_sv_en is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Swedish to English.
### How to use
Here is how to use this model to translate legal text from Swedish to English in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_sv_en"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_sv_en", do_lower_case=False,
skip_special_tokens=True),
device=0
)
sv_text = "Om rättsliga förfaranden inleds rörande omständigheter som ombudsmannen utreder skall han avsluta ärendet."
pipeline([sv_text], max_length=512)
```
## Training data
The legal_t5_small_trans_sv_en model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_sv_en | 52.025|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/code_trans_t5_small_code_documentation_generation_php_multitask | SEBIS | 2021-06-23T10:07:44Z | 13 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "public static function update ( $ table ) { if ( ! is_array ( $ table ) ) { $ table = json_decode ( $ table , true ) ; } if ( ! SchemaManager :: tableExists ( $ table [ 'oldName' ] ) ) { throw SchemaException :: tableDoesNotExist ( $ table [ 'oldName' ] ) ; } $ updater = new self ( $ table ) ; $ updater -> updateTable ( ) ; }"
---
# CodeTrans model for code documentation generation php
Pretrained model on programming language php using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized php code functions: it works best with tokenized php functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets.
## Intended uses & limitations
The model could be used to generate the description for the php function or be fine-tuned on other php code tasks. It can be used on unparsed and untokenized php code. However, if the php code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate php function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_php_multitask"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_php_multitask", skip_special_tokens=True),
device=0
)
tokenized_code = "public static function update ( $ table ) { if ( ! is_array ( $ table ) ) { $ table = json_decode ( $ table , true ) ; } if ( ! SchemaManager :: tableExists ( $ table [ 'oldName' ] ) ) { throw SchemaException :: tableDoesNotExist ( $ table [ 'oldName' ] ) ; } $ updater = new self ( $ table ) ; $ updater -> updateTable ( ) ; }"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/pre-training/function%20documentation%20generation/php/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 420,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
Test results :
| Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
| -------------------- | :------------: | :------------: | :------------: | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 17.31 | 16.65 | 16.89 | 23.05 | 9.19 | 13.7 |
| CodeTrans-ST-Base | 16.86 | 17.17 | 17.16 | 22.98 | 8.23 | 13.17 |
| CodeTrans-TF-Small | 19.93 | 19.48 | 18.88 | 25.35 | 13.15 | 17.23 |
| CodeTrans-TF-Base | 20.26 | 20.19 | 19.50 | 25.84 | 14.07 | 18.25 |
| CodeTrans-TF-Large | 20.35 | 20.06 | **19.54** | 26.18 | 14.94 | **18.98** |
| CodeTrans-MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
| CodeTrans-MT-Base | **20.39** | 21.22 | 19.43 | **26.23** | **15.26** | 16.11 |
| CodeTrans-MT-Large | 20.18 | **21.87** | 19.38 | 26.08 | 15.00 | 16.23 |
| CodeTrans-MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
| CodeTrans-MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | 18.62 |
| CodeTrans-MT-TF-Large | 18.94 | 21.42 | 18.77 | 26.20 | 14.19 | 18.83 |
| State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/legal_t5_small_trans_sv_de | SEBIS | 2021-06-23T10:06:54Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Swedish Deustch model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Swedish Deustch
tags:
- translation Swedish Deustch model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "b) Bekämpning av skadegörare inom skogsbruket."
---
# legal_t5_small_trans_sv_de model
Model on translating legal text from Swedish to Deustch. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_sv_de is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Swedish to Deustch.
### How to use
Here is how to use this model to translate legal text from Swedish to Deustch in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_sv_de"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_sv_de", do_lower_case=False,
skip_special_tokens=True),
device=0
)
sv_text = "b) Bekämpning av skadegörare inom skogsbruket."
pipeline([sv_text], max_length=512)
```
## Training data
The legal_t5_small_trans_sv_de model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_sv_de | 40.264|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/code_trans_t5_small_code_documentation_generation_javascript_multitask_finetune | SEBIS | 2021-06-23T10:05:49Z | 9 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "function isStandardBrowserEnv ( ) { if ( typeof navigator !== 'undefined' && ( navigator . product === 'ReactNative' || navigator . product === 'NativeScript' || navigator . product === 'NS' ) ) { return false ; } return ( typeof window !== 'undefined' && typeof document !== 'undefined' ) ; }"
---
# CodeTrans model for code documentation generation javascript
Pretrained model on programming language javascript using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized javascript code functions: it works best with tokenized javascript functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets. It is then fine-tuned on the code documentation generation task for the javascript function/method.
## Intended uses & limitations
The model could be used to generate the description for the javascript function or be fine-tuned on other javascript code tasks. It can be used on unparsed and untokenized javascript code. However, if the javascript code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate javascript function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_javascript_multitask_finetune"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_javascript_multitask_finetune", skip_special_tokens=True),
device=0
)
tokenized_code = "function isStandardBrowserEnv ( ) { if ( typeof navigator !== 'undefined' && ( navigator . product === 'ReactNative' || navigator . product === 'NativeScript' || navigator . product === 'NS' ) ) { return false ; } return ( typeof window !== 'undefined' && typeof document !== 'undefined' ) ; }"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/fine-tuning/function%20documentation%20generation/javascript/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for half million steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Fine-tuning
This model was then fine-tuned on a single TPU Pod V2-8 for 32,000 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing javascript code.
## Evaluation results
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
| -------------------- | :------------: | :------------: | :------------: | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 17.31 | 16.65 | 16.89 | 23.05 | 9.19 | 13.7 |
| CodeTrans-ST-Base | 16.86 | 17.17 | 17.16 | 22.98 | 8.23 | 13.17 |
| CodeTrans-TF-Small | 19.93 | 19.48 | 18.88 | 25.35 | 13.15 | 17.23 |
| CodeTrans-TF-Base | 20.26 | 20.19 | 19.50 | 25.84 | 14.07 | 18.25 |
| CodeTrans-TF-Large | 20.35 | 20.06 | **19.54** | 26.18 | 14.94 | **18.98** |
| CodeTrans-MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
| CodeTrans-MT-Base | **20.39** | 21.22 | 19.43 | **26.23** | **15.26** | 16.11 |
| CodeTrans-MT-Large | 20.18 | **21.87** | 19.38 | 26.08 | 15.00 | 16.23 |
| CodeTrans-MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
| CodeTrans-MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | 18.62 |
| CodeTrans-MT-TF-Large | 18.94 | 21.42 | 18.77 | 26.20 | 14.19 | 18.83 |
| State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_code_documentation_generation_javascript_multitask | SEBIS | 2021-06-23T10:04:56Z | 13 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "function isStandardBrowserEnv ( ) { if ( typeof navigator !== 'undefined' && ( navigator . product === 'ReactNative' || navigator . product === 'NativeScript' || navigator . product === 'NS' ) ) { return false ; } return ( typeof window !== 'undefined' && typeof document !== 'undefined' ) ; }"
---
# CodeTrans model for code documentation generation javascript
Pretrained model on programming language javascript using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized javascript code functions: it works best with tokenized javascript functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets.
## Intended uses & limitations
The model could be used to generate the description for the javascript function or be fine-tuned on other javascript code tasks. It can be used on unparsed and untokenized javascript code. However, if the javascript code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate javascript function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_javascript_multitask"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_javascript_multitask", skip_special_tokens=True),
device=0
)
tokenized_code = "function isStandardBrowserEnv ( ) { if ( typeof navigator !== 'undefined' && ( navigator . product === 'ReactNative' || navigator . product === 'NativeScript' || navigator . product === 'NS' ) ) { return false ; } return ( typeof window !== 'undefined' && typeof document !== 'undefined' ) ; }"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/pre-training/function%20documentation%20generation/javascript/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 500,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
## Evaluation results
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
| -------------------- | :------------: | :------------: | :------------: | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 17.31 | 16.65 | 16.89 | 23.05 | 9.19 | 13.7 |
| CodeTrans-ST-Base | 16.86 | 17.17 | 17.16 | 22.98 | 8.23 | 13.17 |
| CodeTrans-TF-Small | 19.93 | 19.48 | 18.88 | 25.35 | 13.15 | 17.23 |
| CodeTrans-TF-Base | 20.26 | 20.19 | 19.50 | 25.84 | 14.07 | 18.25 |
| CodeTrans-TF-Large | 20.35 | 20.06 | **19.54** | 26.18 | 14.94 | **18.98** |
| CodeTrans-MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
| CodeTrans-MT-Base | **20.39** | 21.22 | 19.43 | **26.23** | **15.26** | 16.11 |
| CodeTrans-MT-Large | 20.18 | **21.87** | 19.38 | 26.08 | 15.00 | 16.23 |
| CodeTrans-MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
| CodeTrans-MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | 18.62 |
| CodeTrans-MT-TF-Large | 18.94 | 21.42 | 18.77 | 26.20 | 14.19 | 18.83 |
| State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/legal_t5_small_trans_it_sv_small_finetuned | SEBIS | 2021-06-23T10:04:50Z | 4 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Italian Swedish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Italian Swedish
tags:
- translation Italian Swedish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Cooperazione rafforzata Annuncio in Aula"
---
# legal_t5_small_trans_it_sv_small_finetuned model
Model on translating legal text from Italian to Swedish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is first pretrained all the translation data over some unsupervised task. Then the model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_it_sv_small_finetuned is initially pretrained on unsupervised task with the all of the data of the training set. The unsupervised task was "masked language modelling". legal_t5_small_trans_it_sv_small_finetuned is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Italian to Swedish.
### How to use
Here is how to use this model to translate legal text from Italian to Swedish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_it_sv_small_finetuned"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_it_sv", do_lower_case=False,
skip_special_tokens=True),
device=0
)
it_text = "Cooperazione rafforzata Annuncio in Aula"
pipeline([it_text], max_length=512)
```
## Training data
The legal_t5_small_trans_it_sv_small_finetuned (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 8 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
The pre-training data was the combined data from all the 42 language pairs. The task for the model was to predict the portions of a sentence which were masked randomly.
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_it_sv_small_finetuned | 41.243|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_it_sv | SEBIS | 2021-06-23T10:04:14Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Italian Swedish model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Italian Swedish
tags:
- translation Italian Swedish model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "K. considerando che, come avviene con tutti i sistemi di sanità elettronica, la progettazione, lo sviluppo e l’attuazione di sistemi abilitati alla tecnologia RFID presuppongono il coinvolgimento diretto dei professionisti sanitari, dei pazienti e delle commissioni competenti (per esempio, sulla protezione dei dati e sull’etica),"
---
# legal_t5_small_trans_it_sv model
Model on translating legal text from Italian to Swedish. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_it_sv is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Italian to Swedish.
### How to use
Here is how to use this model to translate legal text from Italian to Swedish in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_it_sv"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_it_sv", do_lower_case=False,
skip_special_tokens=True),
device=0
)
it_text = "K. considerando che, come avviene con tutti i sistemi di sanità elettronica, la progettazione, lo sviluppo e l’attuazione di sistemi abilitati alla tecnologia RFID presuppongono il coinvolgimento diretto dei professionisti sanitari, dei pazienti e delle commissioni competenti (per esempio, sulla protezione dei dati e sull’etica),"
pipeline([it_text], max_length=512)
```
## Training data
The legal_t5_small_trans_it_sv model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_it_sv | 41.508|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_it_fr_small_finetuned | SEBIS | 2021-06-23T10:03:39Z | 5 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Italian French model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Italian French
tags:
- translation Italian French model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Dichiarazioni del Consiglio e della Commissione"
---
# legal_t5_small_trans_it_fr_small_finetuned model
Model on translating legal text from Italian to French. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is first pretrained all the translation data over some unsupervised task. Then the model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_it_fr_small_finetuned is initially pretrained on unsupervised task with the all of the data of the training set. The unsupervised task was "masked language modelling". legal_t5_small_trans_it_fr_small_finetuned is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Italian to French.
### How to use
Here is how to use this model to translate legal text from Italian to French in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_it_fr_small_finetuned"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_it_fr", do_lower_case=False,
skip_special_tokens=True),
device=0
)
it_text = "Dichiarazioni del Consiglio e della Commissione"
pipeline([it_text], max_length=512)
```
## Training data
The legal_t5_small_trans_it_fr_small_finetuned (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 9 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
The pre-training data was the combined data from all the 42 language pairs. The task for the model was to predict the portions of a sentence which were masked randomly.
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_it_fr_small_finetuned | 50.557|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/legal_t5_small_trans_it_fr | SEBIS | 2021-06-23T10:03:01Z | 3 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Italian French model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Italian French
tags:
- translation Italian French model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Qualora gli emendamenti approvati dal Parlamento abbiano l'effetto di aumentare le spese iscritte nel progetto di bilancio oltre il tasso massimo previsto, la commissione competente per il merito sottopone al Parlamento una proposta intesa a fissare un nuovo tasso massimo in conformità del paragrafo 9, ultimo comma, degli articoli 78 del trattato CECA, 272 del trattato CE e 177 del trattato CEEA."
---
# legal_t5_small_trans_it_fr model
Model on translating legal text from Italian to French. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_it_fr is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Italian to French.
### How to use
Here is how to use this model to translate legal text from Italian to French in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_it_fr"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_it_fr", do_lower_case=False,
skip_special_tokens=True),
device=0
)
it_text = "Qualora gli emendamenti approvati dal Parlamento abbiano l'effetto di aumentare le spese iscritte nel progetto di bilancio oltre il tasso massimo previsto, la commissione competente per il merito sottopone al Parlamento una proposta intesa a fissare un nuovo tasso massimo in conformità del paragrafo 9, ultimo comma, degli articoli 78 del trattato CECA, 272 del trattato CE e 177 del trattato CEEA."
pipeline([it_text], max_length=512)
```
## Training data
The legal_t5_small_trans_it_fr model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_it_fr | 50.559|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/code_trans_t5_small_code_documentation_generation_java_transfer_learning_finetune | SEBIS | 2021-06-23T10:02:12Z | 10 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "public static < T , U > Function < T , U > castFunction ( Class < U > target ) { return new CastToClass < T , U > ( target ) ; }"
---
# CodeTrans model for code documentation generation java
Pretrained model on programming language java using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized java code functions: it works best with tokenized java functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used transfer-learning pre-training on 7 unsupervised datasets in the software development domain. It is then fine-tuned on the code documentation generation task for the java function/method.
## Intended uses & limitations
The model could be used to generate the description for the java function or be fine-tuned on other java code tasks. It can be used on unparsed and untokenized java code. However, if the java code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate java function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_java_transfer_learning_finetune"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_java_transfer_learning_finetune", skip_special_tokens=True),
device=0
)
tokenized_code = "public static < T , U > Function < T , U > castFunction ( Class < U > target ) { return new CastToClass < T , U > ( target ) ; }"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/transfer%20learning%20fine-tuning/function%20documentation%20generation/java/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Transfer-learning Pretraining
The model was trained on a single TPU Pod V3-8 for half million steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Fine-tuning
This model was then fine-tuned on a single TPU Pod V2-8 for 10,000 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing java code.
## Evaluation results
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
| -------------------- | :------------: | :------------: | :------------: | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 17.31 | 16.65 | 16.89 | 23.05 | 9.19 | 13.7 |
| CodeTrans-ST-Base | 16.86 | 17.17 | 17.16 | 22.98 | 8.23 | 13.17 |
| CodeTrans-TF-Small | 19.93 | 19.48 | 18.88 | 25.35 | 13.15 | 17.23 |
| CodeTrans-TF-Base | 20.26 | 20.19 | 19.50 | 25.84 | 14.07 | 18.25 |
| CodeTrans-TF-Large | 20.35 | 20.06 | **19.54** | 26.18 | 14.94 | **18.98** |
| CodeTrans-MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
| CodeTrans-MT-Base | **20.39** | 21.22 | 19.43 | **26.23** | **15.26** | 16.11 |
| CodeTrans-MT-Large | 20.18 | **21.87** | 19.38 | 26.08 | 15.00 | 16.23 |
| CodeTrans-MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
| CodeTrans-MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | 18.62 |
| CodeTrans-MT-TF-Large | 18.94 | 21.42 | 18.77 | 26.20 | 14.19 | 18.83 |
| State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/code_trans_t5_small_code_documentation_generation_java_multitask | SEBIS | 2021-06-23T10:01:01Z | 10 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "public static < T , U > Function < T , U > castFunction ( Class < U > target ) { return new CastToClass < T , U > ( target ) ; }"
---
# CodeTrans model for code documentation generation java
Pretrained model on programming language java using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized java code functions: it works best with tokenized java functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets.
## Intended uses & limitations
The model could be used to generate the description for the java function or be fine-tuned on other java code tasks. It can be used on unparsed and untokenized java code. However, if the java code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate java function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_java_multitask"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_java_multitask", skip_special_tokens=True),
device=0
)
tokenized_code = "public static < T , U > Function < T , U > castFunction ( Class < U > target ) { return new CastToClass < T , U > ( target ) ; }"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/pre-training/function%20documentation%20generation/java/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for 400,000 steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
## Evaluation results
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
| -------------------- | :------------: | :------------: | :------------: | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 17.31 | 16.65 | 16.89 | 23.05 | 9.19 | 13.7 |
| CodeTrans-ST-Base | 16.86 | 17.17 | 17.16 | 22.98 | 8.23 | 13.17 |
| CodeTrans-TF-Small | 19.93 | 19.48 | 18.88 | 25.35 | 13.15 | 17.23 |
| CodeTrans-TF-Base | 20.26 | 20.19 | 19.50 | 25.84 | 14.07 | 18.25 |
| CodeTrans-TF-Large | 20.35 | 20.06 | **19.54** | 26.18 | 14.94 | **18.98** |
| CodeTrans-MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
| CodeTrans-MT-Base | **20.39** | 21.22 | 19.43 | **26.23** | **15.26** | 16.11 |
| CodeTrans-MT-Large | 20.18 | **21.87** | 19.38 | 26.08 | 15.00 | 16.23 |
| CodeTrans-MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
| CodeTrans-MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | 18.62 |
| CodeTrans-MT-TF-Large | 18.94 | 21.42 | 18.77 | 26.20 | 14.19 | 18.83 |
| State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/legal_t5_small_trans_it_en | SEBIS | 2021-06-23T10:00:46Z | 6 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Italian English model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Italian English
tags:
- translation Italian English model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Oggetto: Libertà di culto in Turchia"
---
# legal_t5_small_trans_it_en model
Model on translating legal text from Italian to English. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_it_en is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Italian to English.
### How to use
Here is how to use this model to translate legal text from Italian to English in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_it_en"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_it_en", do_lower_case=False,
skip_special_tokens=True),
device=0
)
it_text = "Oggetto: Libertà di culto in Turchia"
pipeline([it_text], max_length=512)
```
## Training data
The legal_t5_small_trans_it_en model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 5 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_it_en | 50.068|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/code_trans_t5_small_code_documentation_generation_java | SEBIS | 2021-06-23T10:00:26Z | 17 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "public static < T , U > Function < T , U > castFunction ( Class < U > target ) { return new CastToClass < T , U > ( target ) ; }"
---
# CodeTrans model for code documentation generation java
Pretrained model on programming language java using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized java code functions: it works best with tokenized java functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used single-task training on CodeSearchNet Corpus java dataset.
## Intended uses & limitations
The model could be used to generate the description for the java function or be fine-tuned on other java code tasks. It can be used on unparsed and untokenized java code. However, if the java code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate java function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_java"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_java", skip_special_tokens=True),
device=0
)
tokenized_code = "public static < T , U > Function < T , U > castFunction ( Class < U > target ) { return new CastToClass < T , U > ( target ) ; }"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/single%20task/function%20documentation%20generation/java/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Evaluation results
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
| -------------------- | :------------: | :------------: | :------------: | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 17.31 | 16.65 | 16.89 | 23.05 | 9.19 | 13.7 |
| CodeTrans-ST-Base | 16.86 | 17.17 | 17.16 | 22.98 | 8.23 | 13.17 |
| CodeTrans-TF-Small | 19.93 | 19.48 | 18.88 | 25.35 | 13.15 | 17.23 |
| CodeTrans-TF-Base | 20.26 | 20.19 | 19.50 | 25.84 | 14.07 | 18.25 |
| CodeTrans-TF-Large | 20.35 | 20.06 | **19.54** | 26.18 | 14.94 | **18.98** |
| CodeTrans-MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
| CodeTrans-MT-Base | **20.39** | 21.22 | 19.43 | **26.23** | **15.26** | 16.11 |
| CodeTrans-MT-Large | 20.18 | **21.87** | 19.38 | 26.08 | 15.00 | 16.23 |
| CodeTrans-MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
| CodeTrans-MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | 18.62 |
| CodeTrans-MT-TF-Large | 18.94 | 21.42 | 18.77 | 26.20 | 14.19 | 18.83 |
| State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
SEBIS/legal_t5_small_trans_it_de_small_finetuned | SEBIS | 2021-06-23T10:00:06Z | 9 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"text2text-generation",
"translation Italian Deustch model",
"autotrain_compatible",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | text2text-generation | 2022-03-02T23:29:04Z |
---
language: Italian Deustch
tags:
- translation Italian Deustch model
datasets:
- dcep europarl jrc-acquis
widget:
- text: "Interventi sulla votazione:"
---
# legal_t5_small_trans_it_de_small_finetuned model
Model on translating legal text from Italian to Deustch. It was first released in
[this repository](https://github.com/agemagician/LegalTrans). This model is first pretrained all the translation data over some unsupervised task. Then the model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
## Model description
legal_t5_small_trans_it_de_small_finetuned is initially pretrained on unsupervised task with the all of the data of the training set. The unsupervised task was "masked language modelling". legal_t5_small_trans_it_de_small_finetuned is based on the `t5-small` model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using `dmodel = 512`, `dff = 2,048`, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
## Intended uses & limitations
The model could be used for translation of legal texts from Italian to Deustch.
### How to use
Here is how to use this model to translate legal text from Italian to Deustch in PyTorch:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
pipeline = TranslationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_it_de_small_finetuned"),
tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_it_de", do_lower_case=False,
skip_special_tokens=True),
device=0
)
it_text = "Interventi sulla votazione:"
pipeline([it_text], max_length=512)
```
## Training data
The legal_t5_small_trans_it_de_small_finetuned (the supervised task which involved only the corresponding langauge pair and as well as unsupervised task where all of the data of all language pairs were available) model was trained on [JRC-ACQUIS](https://wt-public.emm4u.eu/Acquis/index_2.2.html), [EUROPARL](https://www.statmt.org/europarl/), and [DCEP](https://ec.europa.eu/jrc/en/language-technologies/dcep) dataset consisting of 8 Million parallel texts.
## Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
### Pretraining
The pre-training data was the combined data from all the 42 language pairs. The task for the model was to predict the portions of a sentence which were masked randomly.
## Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|:-----:|:-----:|
| legal_t5_small_trans_it_de_small_finetuned | 40.524|
### BibTeX entry and citation info
> Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)
|
SEBIS/code_trans_t5_small_code_documentation_generation_go_multitask_finetune | SEBIS | 2021-06-23T09:59:15Z | 13 | 0 | transformers | [
"transformers",
"pytorch",
"jax",
"t5",
"feature-extraction",
"summarization",
"text-generation-inference",
"endpoints_compatible",
"region:us"
] | summarization | 2022-03-02T23:29:04Z | ---
tags:
- summarization
widget:
- text: "func ( pr * Progress ) needSnapshotAbort ( ) bool { return pr . State == ProgressStateSnapshot && pr . Match >= pr . PendingSnapshot }"
---
# CodeTrans model for code documentation generation go
Pretrained model on programming language go using the t5 small model architecture. It was first released in
[this repository](https://github.com/agemagician/CodeTrans). This model is trained on tokenized go code functions: it works best with tokenized go functions.
## Model description
This CodeTrans model is based on the `t5-small` model. It has its own SentencePiece vocabulary model. It used multi-task training on 13 supervised tasks in the software development domain and 7 unsupervised datasets. It is then fine-tuned on the code documentation generation task for the go function/method.
## Intended uses & limitations
The model could be used to generate the description for the go function or be fine-tuned on other go code tasks. It can be used on unparsed and untokenized go code. However, if the go code is tokenized, the performance should be better.
### How to use
Here is how to use this model to generate go function documentation using Transformers SummarizationPipeline:
```python
from transformers import AutoTokenizer, AutoModelWithLMHead, SummarizationPipeline
pipeline = SummarizationPipeline(
model=AutoModelWithLMHead.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_go_multitask_finetune"),
tokenizer=AutoTokenizer.from_pretrained("SEBIS/code_trans_t5_small_code_documentation_generation_go_multitask_finetune", skip_special_tokens=True),
device=0
)
tokenized_code = "func ( pr * Progress ) needSnapshotAbort ( ) bool { return pr . State == ProgressStateSnapshot && pr . Match >= pr . PendingSnapshot }"
pipeline([tokenized_code])
```
Run this example in [colab notebook](https://github.com/agemagician/CodeTrans/blob/main/prediction/multitask/fine-tuning/function%20documentation%20generation/go/small_model.ipynb).
## Training data
The supervised training tasks datasets can be downloaded on [Link](https://www.dropbox.com/sh/488bq2of10r4wvw/AACs5CGIQuwtsD7j_Ls_JAORa/finetuning_dataset?dl=0&subfolder_nav_tracking=1)
## Training procedure
### Multi-task Pretraining
The model was trained on a single TPU Pod V3-8 for half million steps in total, using sequence length 512 (batch size 4096).
It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture.
The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
### Fine-tuning
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 go code.
## Evaluation results
For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
Test results :
| Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
| -------------------- | :------------: | :------------: | :------------: | :------------: | :------------: | :------------: |
| CodeTrans-ST-Small | 17.31 | 16.65 | 16.89 | 23.05 | 9.19 | 13.7 |
| CodeTrans-ST-Base | 16.86 | 17.17 | 17.16 | 22.98 | 8.23 | 13.17 |
| CodeTrans-TF-Small | 19.93 | 19.48 | 18.88 | 25.35 | 13.15 | 17.23 |
| CodeTrans-TF-Base | 20.26 | 20.19 | 19.50 | 25.84 | 14.07 | 18.25 |
| CodeTrans-TF-Large | 20.35 | 20.06 | **19.54** | 26.18 | 14.94 | **18.98** |
| CodeTrans-MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
| CodeTrans-MT-Base | **20.39** | 21.22 | 19.43 | **26.23** | **15.26** | 16.11 |
| CodeTrans-MT-Large | 20.18 | **21.87** | 19.38 | 26.08 | 15.00 | 16.23 |
| CodeTrans-MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
| CodeTrans-MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | 18.62 |
| CodeTrans-MT-TF-Large | 18.94 | 21.42 | 18.77 | 26.20 | 14.19 | 18.83 |
| State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
> Created by [Ahmed Elnaggar](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/) and Wei Ding | [LinkedIn](https://www.linkedin.com/in/wei-ding-92561270/)
|
Subsets and Splits