Update multilingual passage rereanking model card (#6788)

Fix range of possible score, add inference .
This commit is contained in:
Igli Manaj
2020-09-01 23:56:19 +02:00
committed by GitHub
parent 5d820f3ca6
commit f5d69c75f7

View File

@@ -27,13 +27,13 @@ It can be used as an improvement for Elasticsearch Results and boosts the releva
**Architecture:** On top of BERT there is a Densly Connected NN which takes the 768 Dimensional [CLS] Token as input and provides the output ([Arxiv](https://arxiv.org/abs/1901.04085)). **Architecture:** On top of BERT there is a Densly Connected NN which takes the 768 Dimensional [CLS] Token as input and provides the output ([Arxiv](https://arxiv.org/abs/1901.04085)).
**Output:** Just a single value between between 0-1 **Output:** Just a single value between between -10 and 10. Better matching query,passage pairs tend to have a higher a score.
## Intended uses & limitations ## Intended uses & limitations
Both query[1] and passage[2] have to fit in 512 Tokens. Both query[1] and passage[2] have to fit in 512 Tokens.
As you normally want to rerank the first dozens of search results keep in mind the inference time. As you normally want to rerank the first dozens of search results keep in mind the inference time of approximately 300 ms/query.
#### How to use #### How to use
@@ -70,7 +70,7 @@ We see nearly similar performance than the English only Model in the English [Bi
Fine-tuned Models | Dependency | Eval Set | Search Boost<a href='#benchmarks'> | Speed on GPU Fine-tuned Models | Dependency | Eval Set | Search Boost<a href='#benchmarks'> | Speed on GPU
----------------------------------------------------------------------------------- | ---------------------------------------------------------------------------- | ------------------------------------------------------------------ | ----------------------------------------------------- | ---------------------------------- ----------------------------------------------------------------------------------- | ---------------------------------------------------------------------------- | ------------------------------------------------------------------ | ----------------------------------------------------- | ----------------------------------
**`amberoad/Multilingual-uncased-MSMARCO`** (This Model) | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-blue"/> | <a href ='http://www.msmarco.org/'>bing queries</a> | **+61%** <sub><sup>(0.29 vs 0.18)</sup></sub> | - <a href='#footnotes'> **`amberoad/Multilingual-uncased-MSMARCO`** (This Model) | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-blue"/> | <a href ='http://www.msmarco.org/'>bing queries</a> | **+61%** <sub><sup>(0.29 vs 0.18)</sup></sub> | ~300 ms/query <a href='#footnotes'>
`nboost/pt-tinybert-msmarco` | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-red"/> | <a href ='http://www.msmarco.org/'>bing queries</a> | **+45%** <sub><sup>(0.26 vs 0.18)</sup></sub> | ~50ms/query <a href='#footnotes'> `nboost/pt-tinybert-msmarco` | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-red"/> | <a href ='http://www.msmarco.org/'>bing queries</a> | **+45%** <sub><sup>(0.26 vs 0.18)</sup></sub> | ~50ms/query <a href='#footnotes'>
`nboost/pt-bert-base-uncased-msmarco` | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-red"/> | <a href ='http://www.msmarco.org/'>bing queries</a> | **+62%** <sub><sup>(0.29 vs 0.18)</sup></sub> | ~300 ms/query<a href='#footnotes'> `nboost/pt-bert-base-uncased-msmarco` | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-red"/> | <a href ='http://www.msmarco.org/'>bing queries</a> | **+62%** <sub><sup>(0.29 vs 0.18)</sup></sub> | ~300 ms/query<a href='#footnotes'>
`nboost/pt-bert-large-msmarco` | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-red"/> | <a href ='http://www.msmarco.org/'>bing queries</a> | **+77%** <sub><sup>(0.32 vs 0.18)</sup></sub> | - `nboost/pt-bert-large-msmarco` | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-red"/> | <a href ='http://www.msmarco.org/'>bing queries</a> | **+77%** <sub><sup>(0.32 vs 0.18)</sup></sub> | -