Cyrile Delestre
commited on
Commit
·
ef292d4
1
Parent(s):
68e6993
Update README.md
Browse files
README.md
CHANGED
|
@@ -43,24 +43,29 @@ We compare the [DistilCamemBERT](https://huggingface.co/cmarkea/distilcamembert-
|
|
| 43 |
|
| 44 |
| **NLI** | **time (ms)** | **MCC (x100)** |
|
| 45 |
| :--------------: | :-----------: | :------------: |
|
| 46 |
-
| [cmarkea/distilcamembert-base-nli](https://huggingface.co/cmarkea/distilcamembert-base-nli) | **51.35** |
|
| 47 |
| [BaptisteDoyen/camembert-base-xnli](https://huggingface.co/BaptisteDoyen/camembert-base-xnli) | 105.0 | 72.67 |
|
| 48 |
-
| [MoritzLaurer/mDeBERTa-v3-base-mnli-xnli](https://huggingface.co/MoritzLaurer/mDeBERTa-v3-base-mnli-xnli) | 299.18 | 75.15 |
|
|
|
|
|
|
|
|
|
|
| 49 |
|
| 50 |
The main advantage of such modelization is to create a zero-shot classifier allowing text classification without training. This task can be summarized by:
|
| 51 |
$$P(hypothesis=c|premise)=\frac{e^{P(premise=entailment\vert hypothesis\; c)}}{\sum_{i\in\mathcal{C}}e^{P(premise=entailment\vert hypothesis\; i)}}$$
|
| 52 |
|
|
|
|
|
|
|
| 53 |
| **Allociné** | **time (ms)** | **MCC (x100)** |
|
| 54 |
| :--------------: | :-----------: | :------------: |
|
| 55 |
-
| [cmarkea/distilcamembert-base-nli](https://huggingface.co/cmarkea/distilcamembert-base-nli) | **205.54** |
|
| 56 |
-
| [BaptisteDoyen/camembert-base-xnli](https://huggingface.co/BaptisteDoyen/camembert-base-xnli) | 378.39 | 73.74
|
| 57 |
| [MoritzLaurer/mDeBERTa-v3-base-mnli-xnli](https://huggingface.co/MoritzLaurer/mDeBERTa-v3-base-mnli-xnli) | 520.58 | 70.05 |
|
| 58 |
|
| 59 |
| **MLSum** | **time (ms)** | **MCC (x100)** |
|
| 60 |
| :--------------: | :-----------: | :------------: |
|
| 61 |
-
| [cmarkea/distilcamembert-base-nli](https://huggingface.co/cmarkea/distilcamembert-base-nli) | **261.99** |
|
| 62 |
-
| [BaptisteDoyen/camembert-base-xnli](https://huggingface.co/BaptisteDoyen/camembert-base-xnli) | 499.45 | 60.14
|
| 63 |
-
| [MoritzLaurer/mDeBERTa-v3-base-mnli-xnli](https://huggingface.co/MoritzLaurer/mDeBERTa-v3-base-mnli-xnli) | 591.34 | 56.06
|
| 64 |
|
| 65 |
How to use DistilCamemBERT-Sentiment
|
| 66 |
------------------------------------
|
|
|
|
| 43 |
|
| 44 |
| **NLI** | **time (ms)** | **MCC (x100)** |
|
| 45 |
| :--------------: | :-----------: | :------------: |
|
| 46 |
+
| [cmarkea/distilcamembert-base-nli](https://huggingface.co/cmarkea/distilcamembert-base-nli) | **51.35** | 66.24 |
|
| 47 |
| [BaptisteDoyen/camembert-base-xnli](https://huggingface.co/BaptisteDoyen/camembert-base-xnli) | 105.0 | 72.67 |
|
| 48 |
+
| [MoritzLaurer/mDeBERTa-v3-base-mnli-xnli](https://huggingface.co/MoritzLaurer/mDeBERTa-v3-base-mnli-xnli) | 299.18 | **75.15** |
|
| 49 |
+
|
| 50 |
+
Zero-shot classification
|
| 51 |
+
------------------------
|
| 52 |
|
| 53 |
The main advantage of such modelization is to create a zero-shot classifier allowing text classification without training. This task can be summarized by:
|
| 54 |
$$P(hypothesis=c|premise)=\frac{e^{P(premise=entailment\vert hypothesis\; c)}}{\sum_{i\in\mathcal{C}}e^{P(premise=entailment\vert hypothesis\; i)}}$$
|
| 55 |
|
| 56 |
+
For this part, we use 2 datasets, the first one: [allocine](https://huggingface.co/datasets/allocine) used for training the sentiment analysis models. Is composed of 2 classes: "positif" and "négatif" appreciation of movies reviews. Here we use "Ce commentaire est {}." as the hypothesis template and "positif" and "négatif" as candidate labels.
|
| 57 |
+
|
| 58 |
| **Allociné** | **time (ms)** | **MCC (x100)** |
|
| 59 |
| :--------------: | :-----------: | :------------: |
|
| 60 |
+
| [cmarkea/distilcamembert-base-nli](https://huggingface.co/cmarkea/distilcamembert-base-nli) | **205.54** | 63.71 |
|
| 61 |
+
| [BaptisteDoyen/camembert-base-xnli](https://huggingface.co/BaptisteDoyen/camembert-base-xnli) | 378.39 | **73.74** |
|
| 62 |
| [MoritzLaurer/mDeBERTa-v3-base-mnli-xnli](https://huggingface.co/MoritzLaurer/mDeBERTa-v3-base-mnli-xnli) | 520.58 | 70.05 |
|
| 63 |
|
| 64 |
| **MLSum** | **time (ms)** | **MCC (x100)** |
|
| 65 |
| :--------------: | :-----------: | :------------: |
|
| 66 |
+
| [cmarkea/distilcamembert-base-nli](https://huggingface.co/cmarkea/distilcamembert-base-nli) | **261.99** | 60.12 |
|
| 67 |
+
| [BaptisteDoyen/camembert-base-xnli](https://huggingface.co/BaptisteDoyen/camembert-base-xnli) | 499.45 | **60.14** |
|
| 68 |
+
| [MoritzLaurer/mDeBERTa-v3-base-mnli-xnli](https://huggingface.co/MoritzLaurer/mDeBERTa-v3-base-mnli-xnli) | 591.34 | 56.06 |
|
| 69 |
|
| 70 |
How to use DistilCamemBERT-Sentiment
|
| 71 |
------------------------------------
|