Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,57 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: mit
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
language:
|
| 4 |
+
- fr
|
| 5 |
+
library_name: transformers
|
| 6 |
+
tags:
|
| 7 |
+
- linformer
|
| 8 |
+
- legal
|
| 9 |
+
- medical
|
| 10 |
+
- RoBERTa
|
| 11 |
+
- pytorch
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
# Jargon-general-base
|
| 15 |
+
|
| 16 |
+
[Jargon](https://hal.science/hal-04535557/file/FB2_domaines_specialises_LREC_COLING24.pdf) is an efficient transformer encoder LM for French, combining the LinFormer attention mechanism with the RoBERTa model architecture.
|
| 17 |
+
|
| 18 |
+
Jargon is available in several versions with different context sizes and types of pre-training corpora.
|
| 19 |
+
|
| 20 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
| 21 |
+
|
| 22 |
+
<!-- This modelcard aims to be a base template for new models. It has been generated using [this raw template](https://github.com/huggingface/huggingface_hub/blob/main/src/huggingface_hub/templates/modelcard_template.md?plain=1).
|
| 23 |
+
-->
|
| 24 |
+
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
## Using Jargon models with HuggingFace transformers
|
| 28 |
+
|
| 29 |
+
You can get started with `jargon-general-base` using the code snippet below:
|
| 30 |
+
|
| 31 |
+
```python
|
| 32 |
+
from transformers import AutoModelForMaskedLM, AutoTokenizer, pipeline
|
| 33 |
+
|
| 34 |
+
tokenizer = AutoTokenizer.from_pretrained("PantagrueLLM/jargon-general-base", trust_remote_code=True)
|
| 35 |
+
model = AutoModelForMaskedLM.from_pretrained("PantagrueLLM/jargon-general-base", trust_remote_code=True)
|
| 36 |
+
|
| 37 |
+
jargon_maskfiller = pipeline("fill-mask", model=model, tokenizer=tokenizer)
|
| 38 |
+
output = jargon_maskfiller("Il est allé au <mask> hier")
|
| 39 |
+
```
|
| 40 |
+
|
| 41 |
+
- **Funded by**
|
| 42 |
+
- GENCI-IDRIS (Grant 2022 A0131013801)
|
| 43 |
+
- French National Research Agency: Pantagruel grant ANR-23-IAS1-0001
|
| 44 |
+
- MIAI@Grenoble Alpes ANR-19-P3IA-0003
|
| 45 |
+
- PROPICTO ANR-20-CE93-0005
|
| 46 |
+
- Lawbot ANR-20-CE38-0013
|
| 47 |
+
- Swiss National Science Foundation (grant PROPICTO N°197864)
|
| 48 |
+
<!-- - **Shared by [optional]:** [More Information Needed] -->
|
| 49 |
+
<!-- - **Model type:** [More Information Needed] -->
|
| 50 |
+
- **Language(s):** French
|
| 51 |
+
- **License:** MIT
|
| 52 |
+
- **Developed by:** Vincent Segonne
|
| 53 |
+
<!-- - **Finetuned from model [optional]:** [More Information Needed] -->
|
| 54 |
+
<!--
|
| 55 |
+
### Model Sources [optional]
|
| 56 |
+
|
| 57 |
+
<!-- Provide the basic links for the model. -->
|