anakin87 commited on
Commit
3b0238f
β€’
1 Parent(s): f57c1d9

update haystack and streamlit

Browse files
Files changed (3) hide show
  1. README.md +10 -8
  2. app_utils/config.py +0 -2
  3. requirements.txt +2 -2
README.md CHANGED
@@ -4,7 +4,7 @@ emoji: 🎸
4
  colorFrom: purple
5
  colorTo: blue
6
  sdk: streamlit
7
- sdk_version: 1.10.0
8
  app_file: Rock_fact_checker.py
9
  pinned: true
10
  models: [sentence-transformers/msmarco-distilbert-base-tas-b, microsoft/deberta-v2-xlarge-mnli]
@@ -15,13 +15,15 @@ license: apache-2.0
15
 
16
  ## *Fact checking baseline combining dense retrieval and textual entailment*
17
 
18
- - [Idea πŸ’‘](#idea)
19
- - [System description πŸͺ„](#system-description)
20
- - [Indexing pipeline](#indexing-pipeline)
21
- - [Search pipeline](#search-pipeline)
22
- - [Limits and possible improvements ✨](#limits-and-possible-improvements)
23
- - [Repository structure πŸ“](#repository-structure)
24
- - [Installation πŸ’»](#installation)
 
 
25
 
26
  ### Idea
27
  πŸ’‘ This project aims to show that a *naive and simple baseline* for fact checking can be built by combining dense retrieval and a textual entailment task.
 
4
  colorFrom: purple
5
  colorTo: blue
6
  sdk: streamlit
7
+ sdk_version: 1.15.0
8
  app_file: Rock_fact_checker.py
9
  pinned: true
10
  models: [sentence-transformers/msmarco-distilbert-base-tas-b, microsoft/deberta-v2-xlarge-mnli]
 
15
 
16
  ## *Fact checking baseline combining dense retrieval and textual entailment*
17
 
18
+ - [Fact Checking 🎸 Rocks!   ](#fact-checking--rocks---)
19
+ - [*Fact checking baseline combining dense retrieval and textual entailment*](#fact-checking-baseline-combining-dense-retrieval-and-textual-entailment)
20
+ - [Idea](#idea)
21
+ - [System description](#system-description)
22
+ - [Indexing pipeline](#indexing-pipeline)
23
+ - [Search pipeline](#search-pipeline)
24
+ - [Limits and possible improvements](#limits-and-possible-improvements)
25
+ - [Repository structure](#repository-structure)
26
+ - [Installation](#installation)
27
 
28
  ### Idea
29
  πŸ’‘ This project aims to show that a *naive and simple baseline* for fact checking can be built by combining dense retrieval and a textual entailment task.
app_utils/config.py CHANGED
@@ -1,5 +1,3 @@
1
- from curses.ascii import NL
2
- from logging import exception
3
  import streamlit as st
4
 
5
  INDEX_DIR = "data/index"
 
 
 
1
  import streamlit as st
2
 
3
  INDEX_DIR = "data/index"
requirements.txt CHANGED
@@ -1,3 +1,3 @@
1
- farm-haystack[faiss]==1.7.1
2
  plotly==5.10.0
3
- streamlit==1.12.0
 
1
+ farm-haystack[faiss]==1.11.1
2
  plotly==5.10.0
3
+ streamlit==1.15.0