File size: 4,387 Bytes
7f46a81
 
8d7a085
 
7f46a81
8a46321
 
7f46a81
8a46321
f26592e
7f46a81
8a46321
bbb0dfd
8a46321
 
 
7f46a81
 
8d7a085
8a46321
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bbb0dfd
8a46321
 
 
bbb0dfd
 
8a46321
 
 
 
 
 
7f46a81
e63fa66
8d7a085
 
 
 
 
8a46321
 
 
 
 
 
 
 
8d7a085
 
 
 
bbb0dfd
8d7a085
8a46321
 
 
 
 
bbb0dfd
8a46321
 
55fd9a2
7f46a81
 
8a46321
 
7f46a81
 
 
7ff5239
8a46321
 
39e2176
8a46321
39e2176
8d7a085
 
bbb0dfd
8d7a085
8a46321
 
 
 
 
bbb0dfd
8a46321
 
55fd9a2
8d7a085
 
8a46321
 
 
7f46a81
8d7a085
 
8a46321
8d7a085
8a46321
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
import requests
import json


class VectaraQuery():
    def __init__(self, api_key: str, corpus_keys: list[str], prompt_name: str = None):
        self.corpus_keys = corpus_keys
        self.api_key = api_key
        self.prompt_name = prompt_name if prompt_name else "vectara-summary-ext-24-05-sml"
        self.conv_id = None

    
    def get_body(self, query_str: str, response_lang: str, stream: False):
        corpora_list = [{
                'corpus_key': corpus_key, 'lexical_interpolation': 0.005
            } for corpus_key in self.corpus_keys
        ]

        return {
            'query': query_str,
            'search':
            {
                'corpora': corpora_list,
                'offset': 0,
                'limit': 50,
                'context_configuration':
                {
                    'sentences_before': 2,
                    'sentences_after': 2,
                    'start_tag': "%START_SNIPPET%",
                    'end_tag': "%END_SNIPPET%",
                },
                'reranker':
                {
                    'type': 'customer_reranker',
		        'reranker_id': 'rnk_272725719'
                },
            },
            'generation':
            {
                'prompt_name': self.prompt_name,
                'max_used_search_results': 10,
                'response_language': response_lang,
                'citations':
                {
                    'style': 'none'
                },
                'enable_factual_consistency_score': False
            },
            'chat':
            {
                'store': True
            },
            'stream_response': stream
        }
    

    def get_headers(self):
        return {
            "Content-Type": "application/json",
            "Accept": "application/json",
            "x-api-key": self.api_key,
            "grpc-timeout": "60S"
        }
    
    def get_stream_headers(self):
        return {
            "Content-Type": "application/json",
            "Accept": "text/event-stream",
            "x-api-key": self.api_key,
            "grpc-timeout": "60S"
        }

    def submit_query(self, query_str: str, language: str):

        if self.conv_id:
            endpoint = f"https://api.vectara.io/v2/chats/{self.conv_id}/turns"
        else:
            endpoint = "https://api.vectara.io/v2/chats"

        body = self.get_body(query_str, language, stream=False)

        response = requests.post(endpoint, data=json.dumps(body), verify=True, headers=self.get_headers())

        if response.status_code != 200:
            print(f"Query failed with code {response.status_code}, reason {response.reason}, text {response.text}")
            if response.status_code == 429:
                return "Sorry, Vectara chat turns exceeds plan limit."
            return "Sorry, something went wrong in my brain. Please try again later."

        res = response.json()

        if self.conv_id is None:
            self.conv_id = res['chat_id']

        summary = res['answer']
        
        return summary

    def submit_query_streaming(self, query_str: str, language: str):

        if self.conv_id:
            endpoint = f"https://api.vectara.io/v2/chats/{self.conv_id}/turns"
        else:
            endpoint = "https://api.vectara.io/v2/chats"

        body = self.get_body(query_str, language, stream=True)

        response = requests.post(endpoint, data=json.dumps(body), verify=True, headers=self.get_stream_headers(), stream=True) 

        if response.status_code != 200:
            print(f"Query failed with code {response.status_code}, reason {response.reason}, text {response.text}")
            if response.status_code == 429:
                return "Sorry, Vectara chat turns exceeds plan limit."
            return "Sorry, something went wrong in my brain. Please try again later."        

        chunks = []
        for line in response.iter_lines():
            line = line.decode('utf-8')
            if line:  # filter out keep-alive new lines
                key, value = line.split(':', 1)
                if key == 'data':
                    line = json.loads(value)
                    if line['type'] == 'generation_chunk':
                        chunk = line['generation_chunk']
                        chunks.append(chunk)
                        yield chunk

        return ''.join(chunks)