added telegram bot
Browse files- Makefile +3 -0
- app_modules/llm_inference.py +45 -52
- app_modules/llm_loader.py +37 -21
- requirements.txt +1 -0
- server.py +45 -18
- telegram_bot.py +93 -0
Makefile
CHANGED
@@ -15,6 +15,9 @@ test:
|
|
15 |
chat:
|
16 |
python test.py chat
|
17 |
|
|
|
|
|
|
|
18 |
ingest:
|
19 |
python ingest.py
|
20 |
|
|
|
15 |
chat:
|
16 |
python test.py chat
|
17 |
|
18 |
+
tele:
|
19 |
+
python telegram_bot.py
|
20 |
+
|
21 |
ingest:
|
22 |
python ingest.py
|
23 |
|
app_modules/llm_inference.py
CHANGED
@@ -38,71 +38,64 @@ class LLMInference(metaclass=abc.ABCMeta):
|
|
38 |
self, inputs, streaming_handler, q: Queue = None, tracing: bool = False
|
39 |
):
|
40 |
print(inputs)
|
|
|
41 |
|
42 |
-
|
43 |
-
self.llm_loader.streamer, TextIteratorStreamer
|
44 |
-
):
|
45 |
self.llm_loader.streamer.reset(q)
|
46 |
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
|
|
|
|
|
|
|
|
53 |
)
|
54 |
-
if streaming_handler is not None
|
55 |
-
else chain(inputs)
|
56 |
-
)
|
57 |
|
58 |
-
|
59 |
-
|
60 |
|
61 |
-
|
62 |
-
|
63 |
-
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
|
68 |
|
69 |
-
|
|
|
|
|
70 |
|
71 |
def _execute_chain(self, chain, inputs, q, sh):
|
72 |
q.put(chain(inputs, callbacks=[sh]))
|
73 |
|
74 |
def _run_chain(self, chain, inputs, streaming_handler):
|
75 |
-
|
76 |
-
try:
|
77 |
-
que = Queue()
|
78 |
|
79 |
-
|
80 |
-
|
81 |
-
|
82 |
-
|
83 |
-
|
84 |
-
|
85 |
-
if self.llm_loader.streamer is not None and isinstance(
|
86 |
-
self.llm_loader.streamer, TextIteratorStreamer
|
87 |
-
):
|
88 |
-
count = (
|
89 |
-
2
|
90 |
-
if "chat_history" in inputs and len(inputs.get("chat_history")) > 0
|
91 |
-
else 1
|
92 |
-
)
|
93 |
|
94 |
-
|
95 |
-
|
96 |
-
|
97 |
-
streaming_handler.on_llm_new_token(token)
|
98 |
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
time.sleep(0.5)
|
104 |
|
105 |
-
|
106 |
-
|
107 |
-
|
108 |
-
|
|
|
|
|
|
|
|
|
|
38 |
self, inputs, streaming_handler, q: Queue = None, tracing: bool = False
|
39 |
):
|
40 |
print(inputs)
|
41 |
+
self.llm_loader.lock.acquire()
|
42 |
|
43 |
+
try:
|
|
|
|
|
44 |
self.llm_loader.streamer.reset(q)
|
45 |
|
46 |
+
chain = self.get_chain(tracing)
|
47 |
+
result = (
|
48 |
+
self._run_chain(
|
49 |
+
chain,
|
50 |
+
inputs,
|
51 |
+
streaming_handler,
|
52 |
+
)
|
53 |
+
if streaming_handler is not None
|
54 |
+
and self.llm_loader.streamer.for_huggingface
|
55 |
+
else chain(inputs)
|
56 |
)
|
|
|
|
|
|
|
57 |
|
58 |
+
if "answer" in result:
|
59 |
+
result["answer"] = remove_extra_spaces(result["answer"])
|
60 |
|
61 |
+
base_url = os.environ.get("PDF_FILE_BASE_URL")
|
62 |
+
if base_url is not None and len(base_url) > 0:
|
63 |
+
documents = result["source_documents"]
|
64 |
+
for doc in documents:
|
65 |
+
source = doc.metadata["source"]
|
66 |
+
title = source.split("/")[-1]
|
67 |
+
doc.metadata["url"] = f"{base_url}{urllib.parse.quote(title)}"
|
68 |
|
69 |
+
return result
|
70 |
+
finally:
|
71 |
+
self.llm_loader.lock.release()
|
72 |
|
73 |
def _execute_chain(self, chain, inputs, q, sh):
|
74 |
q.put(chain(inputs, callbacks=[sh]))
|
75 |
|
76 |
def _run_chain(self, chain, inputs, streaming_handler):
|
77 |
+
que = Queue()
|
|
|
|
|
78 |
|
79 |
+
t = Thread(
|
80 |
+
target=self._execute_chain,
|
81 |
+
args=(chain, inputs, que, streaming_handler),
|
82 |
+
)
|
83 |
+
t.start()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
84 |
|
85 |
+
count = (
|
86 |
+
2 if "chat_history" in inputs and len(inputs.get("chat_history")) > 0 else 1
|
87 |
+
)
|
|
|
88 |
|
89 |
+
while count > 0:
|
90 |
+
try:
|
91 |
+
for token in self.llm_loader.streamer:
|
92 |
+
streaming_handler.on_llm_new_token(token)
|
|
|
93 |
|
94 |
+
self.llm_loader.streamer.reset()
|
95 |
+
count -= 1
|
96 |
+
except Exception:
|
97 |
+
print("nothing generated yet - retry in 0.5s")
|
98 |
+
time.sleep(0.5)
|
99 |
+
|
100 |
+
t.join()
|
101 |
+
return que.get()
|
app_modules/llm_loader.py
CHANGED
@@ -33,18 +33,22 @@ class TextIteratorStreamer(TextStreamer, StreamingStdOutCallbackHandler):
|
|
33 |
tokenizer: "AutoTokenizer",
|
34 |
skip_prompt: bool = False,
|
35 |
timeout: Optional[float] = None,
|
|
|
36 |
**decode_kwargs,
|
37 |
):
|
38 |
super().__init__(tokenizer, skip_prompt, **decode_kwargs)
|
39 |
self.text_queue = Queue()
|
40 |
self.stop_signal = None
|
41 |
self.timeout = timeout
|
|
|
|
|
42 |
|
43 |
def on_finalized_text(self, text: str, stream_end: bool = False):
|
44 |
super().on_finalized_text(text, stream_end=stream_end)
|
45 |
|
46 |
"""Put the new text in the queue. If the stream is ending, also put a stop signal in the queue."""
|
47 |
self.text_queue.put(text, timeout=self.timeout)
|
|
|
48 |
if stream_end:
|
49 |
print("\n")
|
50 |
self.text_queue.put("\n", timeout=self.timeout)
|
@@ -54,12 +58,16 @@ class TextIteratorStreamer(TextStreamer, StreamingStdOutCallbackHandler):
|
|
54 |
sys.stdout.write(token)
|
55 |
sys.stdout.flush()
|
56 |
self.text_queue.put(token, timeout=self.timeout)
|
|
|
57 |
|
58 |
def on_llm_end(self, response: LLMResult, **kwargs: Any) -> None:
|
59 |
print("\n")
|
60 |
self.text_queue.put("\n", timeout=self.timeout)
|
61 |
self.text_queue.put(self.stop_signal, timeout=self.timeout)
|
62 |
|
|
|
|
|
|
|
63 |
def __iter__(self):
|
64 |
return self
|
65 |
|
@@ -88,21 +96,18 @@ class LLMLoader:
|
|
88 |
def __init__(self, llm_model_type, lc_serve: bool = False):
|
89 |
self.llm_model_type = llm_model_type
|
90 |
self.llm = None
|
91 |
-
self.streamer =
|
92 |
self.max_tokens_limit = 2048
|
93 |
self.search_kwargs = {"k": 4}
|
94 |
self.lock = threading.Lock()
|
95 |
|
96 |
-
def
|
97 |
-
self.streamer = (
|
98 |
-
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
)
|
104 |
-
if custom_handler is None
|
105 |
-
else TextStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
|
106 |
)
|
107 |
|
108 |
def init(
|
@@ -179,7 +184,11 @@ class LLMLoader:
|
|
179 |
MODEL_NAME_OR_PATH = os.environ.get("HUGGINGFACE_MODEL_NAME_OR_PATH")
|
180 |
print(f" loading model: {MODEL_NAME_OR_PATH}")
|
181 |
|
182 |
-
hf_auth_token =
|
|
|
|
|
|
|
|
|
183 |
transformers_offline = os.environ.get("TRANSFORMERS_OFFLINE") == "1"
|
184 |
token = (
|
185 |
hf_auth_token
|
@@ -231,7 +240,7 @@ class LLMLoader:
|
|
231 |
)
|
232 |
)
|
233 |
|
234 |
-
self.
|
235 |
|
236 |
task = "text2text-generation" if is_t5 else "text-generation"
|
237 |
|
@@ -343,14 +352,21 @@ class LLMLoader:
|
|
343 |
MODEL_NAME_OR_PATH,
|
344 |
config=config,
|
345 |
trust_remote_code=True,
|
346 |
-
token=token,
|
347 |
)
|
348 |
if is_t5
|
349 |
-
else
|
350 |
-
|
351 |
-
|
352 |
-
|
353 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
354 |
)
|
355 |
)
|
356 |
print(f"Model memory footprint: {model.get_memory_footprint()}")
|
@@ -405,7 +421,7 @@ class LLMLoader:
|
|
405 |
print(f"Model memory footprint: {model.get_memory_footprint()}")
|
406 |
|
407 |
tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neox-20b")
|
408 |
-
self.
|
409 |
|
410 |
# mtp-7b is trained to add "<|endoftext|>" at the end of generations
|
411 |
stop_token_ids = tokenizer.convert_tokens_to_ids(["<|endoftext|>"])
|
@@ -497,7 +513,7 @@ class LLMLoader:
|
|
497 |
print(f"Model memory footprint: {model.get_memory_footprint()}")
|
498 |
|
499 |
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME_OR_PATH)
|
500 |
-
self.
|
501 |
|
502 |
class StopOnTokens(StoppingCriteria):
|
503 |
def __call__(
|
|
|
33 |
tokenizer: "AutoTokenizer",
|
34 |
skip_prompt: bool = False,
|
35 |
timeout: Optional[float] = None,
|
36 |
+
for_huggingface: bool = False,
|
37 |
**decode_kwargs,
|
38 |
):
|
39 |
super().__init__(tokenizer, skip_prompt, **decode_kwargs)
|
40 |
self.text_queue = Queue()
|
41 |
self.stop_signal = None
|
42 |
self.timeout = timeout
|
43 |
+
self.total_tokens = 0
|
44 |
+
self.for_huggingface = for_huggingface
|
45 |
|
46 |
def on_finalized_text(self, text: str, stream_end: bool = False):
|
47 |
super().on_finalized_text(text, stream_end=stream_end)
|
48 |
|
49 |
"""Put the new text in the queue. If the stream is ending, also put a stop signal in the queue."""
|
50 |
self.text_queue.put(text, timeout=self.timeout)
|
51 |
+
self.total_tokens = self.total_tokens + 1
|
52 |
if stream_end:
|
53 |
print("\n")
|
54 |
self.text_queue.put("\n", timeout=self.timeout)
|
|
|
58 |
sys.stdout.write(token)
|
59 |
sys.stdout.flush()
|
60 |
self.text_queue.put(token, timeout=self.timeout)
|
61 |
+
self.total_tokens = self.total_tokens + 1
|
62 |
|
63 |
def on_llm_end(self, response: LLMResult, **kwargs: Any) -> None:
|
64 |
print("\n")
|
65 |
self.text_queue.put("\n", timeout=self.timeout)
|
66 |
self.text_queue.put(self.stop_signal, timeout=self.timeout)
|
67 |
|
68 |
+
def for_huggingface(self) -> bool:
|
69 |
+
return self.tokenizer != ""
|
70 |
+
|
71 |
def __iter__(self):
|
72 |
return self
|
73 |
|
|
|
96 |
def __init__(self, llm_model_type, lc_serve: bool = False):
|
97 |
self.llm_model_type = llm_model_type
|
98 |
self.llm = None
|
99 |
+
self.streamer = TextIteratorStreamer("")
|
100 |
self.max_tokens_limit = 2048
|
101 |
self.search_kwargs = {"k": 4}
|
102 |
self.lock = threading.Lock()
|
103 |
|
104 |
+
def _init_hf_streamer(self, tokenizer):
|
105 |
+
self.streamer = TextIteratorStreamer(
|
106 |
+
tokenizer,
|
107 |
+
timeout=10.0,
|
108 |
+
skip_prompt=True,
|
109 |
+
skip_special_tokens=True,
|
110 |
+
for_huggingface=True,
|
|
|
|
|
|
|
111 |
)
|
112 |
|
113 |
def init(
|
|
|
184 |
MODEL_NAME_OR_PATH = os.environ.get("HUGGINGFACE_MODEL_NAME_OR_PATH")
|
185 |
print(f" loading model: {MODEL_NAME_OR_PATH}")
|
186 |
|
187 |
+
hf_auth_token = (
|
188 |
+
os.environ.get("HUGGINGFACE_AUTH_TOKEN")
|
189 |
+
if "Llama-2" in MODEL_NAME_OR_PATH
|
190 |
+
else None
|
191 |
+
)
|
192 |
transformers_offline = os.environ.get("TRANSFORMERS_OFFLINE") == "1"
|
193 |
token = (
|
194 |
hf_auth_token
|
|
|
240 |
)
|
241 |
)
|
242 |
|
243 |
+
self._init_hf_streamer(tokenizer)
|
244 |
|
245 |
task = "text2text-generation" if is_t5 else "text-generation"
|
246 |
|
|
|
352 |
MODEL_NAME_OR_PATH,
|
353 |
config=config,
|
354 |
trust_remote_code=True,
|
|
|
355 |
)
|
356 |
if is_t5
|
357 |
+
else (
|
358 |
+
AutoModelForCausalLM.from_pretrained(
|
359 |
+
MODEL_NAME_OR_PATH,
|
360 |
+
config=config,
|
361 |
+
trust_remote_code=True,
|
362 |
+
)
|
363 |
+
if token is None
|
364 |
+
else AutoModelForCausalLM.from_pretrained(
|
365 |
+
MODEL_NAME_OR_PATH,
|
366 |
+
config=config,
|
367 |
+
trust_remote_code=True,
|
368 |
+
token=token,
|
369 |
+
)
|
370 |
)
|
371 |
)
|
372 |
print(f"Model memory footprint: {model.get_memory_footprint()}")
|
|
|
421 |
print(f"Model memory footprint: {model.get_memory_footprint()}")
|
422 |
|
423 |
tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neox-20b")
|
424 |
+
self._init_hf_streamer(tokenizer)
|
425 |
|
426 |
# mtp-7b is trained to add "<|endoftext|>" at the end of generations
|
427 |
stop_token_ids = tokenizer.convert_tokens_to_ids(["<|endoftext|>"])
|
|
|
513 |
print(f"Model memory footprint: {model.get_memory_footprint()}")
|
514 |
|
515 |
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME_OR_PATH)
|
516 |
+
self._init_hf_streamer(tokenizer)
|
517 |
|
518 |
class StopOnTokens(StoppingCriteria):
|
519 |
def __call__(
|
requirements.txt
CHANGED
@@ -31,3 +31,4 @@ einops
|
|
31 |
gevent
|
32 |
pydantic >= 1.10.11
|
33 |
pypdf
|
|
|
|
31 |
gevent
|
32 |
pydantic >= 1.10.11
|
33 |
pypdf
|
34 |
+
python-telegram-bot
|
server.py
CHANGED
@@ -11,7 +11,7 @@ from app_modules.init import app_init
|
|
11 |
from app_modules.llm_chat_chain import ChatChain
|
12 |
from app_modules.utils import print_llm_response
|
13 |
|
14 |
-
llm_loader, qa_chain = app_init(
|
15 |
|
16 |
chat_history_enabled = os.environ.get("CHAT_HISTORY_ENABLED") == "true"
|
17 |
|
@@ -26,14 +26,13 @@ class ChatResponse(BaseModel):
|
|
26 |
sourceDocs: Optional[List] = None
|
27 |
|
28 |
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
if uuid is None:
|
37 |
chat_history = []
|
38 |
if chat_history_enabled:
|
39 |
for element in history:
|
@@ -48,21 +47,49 @@ def chat(
|
|
48 |
print(f"Completed in {end - start:.3f}s")
|
49 |
|
50 |
print(f"qa_chain result: {result}")
|
51 |
-
|
52 |
-
|
53 |
-
return json.dumps(resp.dict())
|
54 |
else:
|
55 |
-
if
|
56 |
-
chat = uuid_to_chat_chain_mapping[
|
57 |
else:
|
58 |
chat = ChatChain(llm_loader)
|
59 |
-
uuid_to_chat_chain_mapping[
|
60 |
result = chat.call_chain({"question": question}, streaming_handler)
|
61 |
print(f"chat result: {result}")
|
|
|
|
|
62 |
|
63 |
-
|
64 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
65 |
|
66 |
|
67 |
if __name__ == "__main__":
|
68 |
-
print_llm_response(json.loads(chat("What's deep learning?", [])))
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
from app_modules.llm_chat_chain import ChatChain
|
12 |
from app_modules.utils import print_llm_response
|
13 |
|
14 |
+
llm_loader, qa_chain = app_init(__name__ != "__main__")
|
15 |
|
16 |
chat_history_enabled = os.environ.get("CHAT_HISTORY_ENABLED") == "true"
|
17 |
|
|
|
26 |
sourceDocs: Optional[List] = None
|
27 |
|
28 |
|
29 |
+
def do_chat(
|
30 |
+
question: str,
|
31 |
+
history: Optional[List] = [],
|
32 |
+
chat_id: Optional[str] = None,
|
33 |
+
streaming_handler: any = None,
|
34 |
+
):
|
35 |
+
if chat_id is None:
|
|
|
36 |
chat_history = []
|
37 |
if chat_history_enabled:
|
38 |
for element in history:
|
|
|
47 |
print(f"Completed in {end - start:.3f}s")
|
48 |
|
49 |
print(f"qa_chain result: {result}")
|
50 |
+
return result
|
|
|
|
|
51 |
else:
|
52 |
+
if chat_id in uuid_to_chat_chain_mapping:
|
53 |
+
chat = uuid_to_chat_chain_mapping[chat_id]
|
54 |
else:
|
55 |
chat = ChatChain(llm_loader)
|
56 |
+
uuid_to_chat_chain_mapping[chat_id] = chat
|
57 |
result = chat.call_chain({"question": question}, streaming_handler)
|
58 |
print(f"chat result: {result}")
|
59 |
+
return result
|
60 |
+
|
61 |
|
62 |
+
@serving(websocket=True)
|
63 |
+
def chat(
|
64 |
+
question: str, history: Optional[List] = [], chat_id: Optional[str] = None, **kwargs
|
65 |
+
) -> str:
|
66 |
+
print("question@chat:", question)
|
67 |
+
streaming_handler = kwargs.get("streaming_handler")
|
68 |
+
result = do_chat(question, history, chat_id, streaming_handler)
|
69 |
+
resp = ChatResponse(
|
70 |
+
sourceDocs=result["source_documents"] if chat_id is None else []
|
71 |
+
)
|
72 |
+
return json.dumps(resp.dict())
|
73 |
+
|
74 |
+
|
75 |
+
@serving
|
76 |
+
def chat_sync(
|
77 |
+
question: str, history: Optional[List] = [], chat_id: Optional[str] = None, **kwargs
|
78 |
+
) -> str:
|
79 |
+
print("question@chat_sync:", question)
|
80 |
+
result = do_chat(question, history, chat_id, None)
|
81 |
+
return result["text"]
|
82 |
|
83 |
|
84 |
if __name__ == "__main__":
|
85 |
+
# print_llm_response(json.loads(chat("What's deep learning?", [])))
|
86 |
+
chat_start = timer()
|
87 |
+
chat_sync("What's generative AI?", chat_id="test_user")
|
88 |
+
chat_sync("more on finance", chat_id="test_user")
|
89 |
+
chat_end = timer()
|
90 |
+
total_time = chat_end - chat_start
|
91 |
+
print(f"Total time used: {total_time:.3f} s")
|
92 |
+
print(f"Number of tokens generated: {llm_loader.streamer.total_tokens}")
|
93 |
+
print(
|
94 |
+
f"Average generation speed: {llm_loader.streamer.total_tokens / total_time:.3f} tokens/s"
|
95 |
+
)
|
telegram_bot.py
ADDED
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
import ssl
|
3 |
+
import time
|
4 |
+
from threading import Thread
|
5 |
+
|
6 |
+
import requests
|
7 |
+
from telegram import Update
|
8 |
+
from telegram import __version__ as TG_VER
|
9 |
+
from telegram.ext import (
|
10 |
+
Application,
|
11 |
+
CommandHandler,
|
12 |
+
ContextTypes,
|
13 |
+
MessageHandler,
|
14 |
+
filters,
|
15 |
+
)
|
16 |
+
|
17 |
+
from app_modules.init import *
|
18 |
+
|
19 |
+
ctx = ssl.create_default_context()
|
20 |
+
ctx.set_ciphers("DEFAULT")
|
21 |
+
|
22 |
+
try:
|
23 |
+
from telegram import __version_info__
|
24 |
+
except ImportError:
|
25 |
+
__version_info__ = (0, 0, 0, 0, 0) # type: ignore[assignment]
|
26 |
+
|
27 |
+
if __version_info__ < (20, 0, 0, "alpha", 1):
|
28 |
+
raise RuntimeError(
|
29 |
+
f"This example is not compatible with your current PTB version {TG_VER}. To view the "
|
30 |
+
f"{TG_VER} version of this example, "
|
31 |
+
f"visit https://docs.python-telegram-bot.org/en/v{TG_VER}/examples.html"
|
32 |
+
)
|
33 |
+
|
34 |
+
TOKEN = os.getenv("TELEGRAM_API_TOKEN")
|
35 |
+
ENDPOINT = os.getenv("CHAT_API_URL")
|
36 |
+
|
37 |
+
|
38 |
+
# Define a few command handlers. These usually take the two arguments update and
|
39 |
+
# context.
|
40 |
+
async def start_command(update: Update, context: ContextTypes.DEFAULT_TYPE) -> None:
|
41 |
+
"""Send a message when the command /start is issued."""
|
42 |
+
user = update.effective_user
|
43 |
+
await update.message.reply_html(
|
44 |
+
rf"Hi {user.mention_html()}! You are welcome to ask questions on anything!",
|
45 |
+
)
|
46 |
+
|
47 |
+
|
48 |
+
async def help_command(update: Update, context: ContextTypes.DEFAULT_TYPE) -> None:
|
49 |
+
"""Send a message when the command /help is issued."""
|
50 |
+
await update.message.reply_text("Help!")
|
51 |
+
|
52 |
+
|
53 |
+
async def chat_command(update: Update, context: ContextTypes.DEFAULT_TYPE) -> None:
|
54 |
+
"""Echo the user message."""
|
55 |
+
tic = time.perf_counter()
|
56 |
+
try:
|
57 |
+
message = {
|
58 |
+
"question": update.message.text,
|
59 |
+
"chat_id": update.message.chat.username,
|
60 |
+
}
|
61 |
+
print(message)
|
62 |
+
x = requests.post(ENDPOINT, json=message).json()
|
63 |
+
temp = time.perf_counter()
|
64 |
+
print(f"Received response in {temp - tic:0.4f} seconds")
|
65 |
+
result = x["result"]
|
66 |
+
print(result)
|
67 |
+
await update.message.reply_text(result)
|
68 |
+
toc = time.perf_counter()
|
69 |
+
print(f"Response time in {toc - tic:0.4f} seconds")
|
70 |
+
except Exception as e:
|
71 |
+
print("error", e)
|
72 |
+
|
73 |
+
|
74 |
+
def start_telegram_bot() -> None:
|
75 |
+
"""Start the bot."""
|
76 |
+
print("starting telegram bot ...")
|
77 |
+
# Create the Application and pass it your bot's token.
|
78 |
+
application = Application.builder().token(TOKEN).build()
|
79 |
+
|
80 |
+
# on different commands - answer in Telegram
|
81 |
+
application.add_handler(CommandHandler("start_command", start_command))
|
82 |
+
application.add_handler(CommandHandler("help", help_command))
|
83 |
+
|
84 |
+
# on non command i.e message - chat_command the message on Telegram
|
85 |
+
application.add_handler(
|
86 |
+
MessageHandler(filters.TEXT & ~filters.COMMAND, chat_command)
|
87 |
+
)
|
88 |
+
|
89 |
+
application.run_polling()
|
90 |
+
|
91 |
+
|
92 |
+
if __name__ == "__main__":
|
93 |
+
start_telegram_bot()
|