Hugging Face's logo

Spaces:
rr19tech
/
RetroPosterBoard
Running

App Files Community
rr19tech commited on
Commit
3ff49ac
·
verified ·
1 Parent(s): 9f4f2e5

Updated new article quality check project.

Browse files
Files changed (1) hide show
  1. Articles_Quality_Check.html +160 -0
Articles_Quality_Check.html ADDED
@@ -0,0 +1,160 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ <!DOCTYPE html>
2
+ <html lang="en">
3
+ <head>
4
+ <meta charset="UTF-8">
5
+ <meta name="viewport" content="width=device-width, initial-scale=1.0">
6
+ <title>A Project to quality check the submitted articles</title>
7
+ <style>
8
+ body {
9
+ font-family: Arial, sans-serif;
10
+ line-height: 1.6;
11
+ margin: 20px;
12
+ }
13
+ h1, h2 {
14
+ color: #333;
15
+ }
16
+ h2 {
17
+ margin-top: 30px;
18
+ }
19
+ ul {
20
+ list-style-type: disc;
21
+ margin-left: 20px;
22
+ }
23
+ p {
24
+ margin-bottom: 15px;
25
+ }
26
+
27
+ table{
28
+ border-collapse: collapse;
29
+ width: 95%;
30
+ border: 2px solid #2c3e50;
31
+ }
32
+
33
+ tr{
34
+ border-bottom: 2px solid #b60e0e;
35
+ }
36
+
37
+ td{
38
+ width: 15%;
39
+ vertical-align: top;
40
+ border: 2px solid #3498db;
41
+ }
42
+ </style>
43
+ </head>
44
+ <body>
45
+ <h2>Overview of the project</h2>
46
+ <p>
47
+ In this project we will evaluate the quality of all the articles submitted by the content writers.
48
+ All the articles are fetched based on a time window (last 3 days, last one week etc). Llamaindex workflow will be utilized
49
+ to check each article for its structure, any missing elements are there in the article and give an overall feedback.
50
+ We use a pydantic schema to predefine the structure and llm will be using the sturcture predict to evaluate.
51
+ A final report is created.
52
+ </p>
53
+ <h2>Here is the design, we will examine two approaches</h2>
54
+ <p>
55
+ First approach, each is a seperate task
56
+ <pre>
57
+ +----------------------------------+
58
+
59
+ | User Time Window |
60
+ | (e.g., "Last 24 Hours") |
61
+ +----------------+-----------------+
62
+ |
63
+ v
64
+ +----------------+-----------------+
65
+
66
+ | CMS System |
67
+ | (Fetch Articles in Time Window) |
68
+ +----------------+-----------------+
69
+ |
70
+ v
71
+ +-------------------------------------------------------------------+
72
+
73
+ | LLAMAINDEX WORKFLOW |
74
+ | |
75
+ | +-----------------------+ +--------------------+ |
76
+ | | Start Event | | Router / Dispatch | |
77
+ | | (Incoming CMS Data) +------------->| (Parallel Execution) | |
78
+ | +-----------------------+ +---------+----------+ |
79
+ | | |
80
+ | +-------------------+----------------------+ |
81
+ | | | | |
82
+ | v v v |
83
+ | +--------------+ +--------------+ +-------------------+ |
84
+ | | Task 1: | | Task 2: | | Task 3: | |
85
+ | | Structure | | Wellformed | | Content Quality | |
86
+ | | Analysis | | Validation | | Assessor | |
87
+ | +------+-------+ +------+-------+ +---------+---------+ |
88
+ | | | | |
89
+ | +-------------------+----------------------+ |
90
+ | | (Emit Partial Results) |
91
+ | v |
92
+ | +--------------------+ |
93
+ | | Gather / Step | |
94
+ | | (Collects 3 Tasks) | |
95
+ | +----------+---------+ |
96
+ | | |
97
+ | v |
98
+ | +--------------------+ |
99
+ | | Stop / Report Step| |
100
+ | | (Synthesize Data) | |
101
+ | +----------+---------+ |
102
+ +-----------------------------|-------------------------------------+
103
+ |
104
+ v
105
+ +--------------+-------------------+
106
+
107
+ | Final Quality Report |
108
+ | (Structure, Format & Insights) |
109
+ +----------------------------------+
110
+
111
+ </pre>
112
+ Second approach is all the tasks are clubbed to one single workflow step task.
113
+ <pre>
114
+ +----------------------------------+
115
+
116
+ | User Time Window |
117
+ | (e.g., "Last 24 Hours") |
118
+ +----------------+-----------------+
119
+ |
120
+ v
121
+ +----------------+-----------------+
122
+
123
+ | CMS System |
124
+ | (Fetch Articles in Time Window) |
125
+ +----------------+-----------------+
126
+ |
127
+ v
128
+ +-------------------------------------------------------------------+
129
+
130
+ | SINGLE-STEP STRUCTURED PREDICT |
131
+ | |
132
+ | +-----------------------+ +--------------------+ |
133
+ | | Raw Article Data +------------->| LLM Context | |
134
+ | +-----------------------+ | (Single Payload) | |
135
+ | +---------+----------+ |
136
+ | | |
137
+ | v |
138
+ | +--------------------+ |
139
+ | |llm.structured_predict |
140
+ | +---------+----------+ |
141
+ | | |
142
+ | v |
143
+ | +-----------------+------------------+
144
+
145
+ | | Unified Pydantic Schema Validation|
146
+ | | [Structure + Format + Quality] |
147
+ | +-----------------+------------------+
148
+ +----------------------------------------------------|--------------+
149
+ |
150
+ v
151
+ +-------------------------------------+------------+
152
+
153
+ | Final Article Report |
154
+ | (Instantly validated JSON / Typed Python Object) |
155
+ +--------------------------------------------------+
156
+ </pre>
157
+
158
+ </p>
159
+ </body>
160
+ </html>