Spaces:
Sleeping
Sleeping
add time log and reduce processing time
Browse files
app.py
CHANGED
@@ -2,42 +2,53 @@ from fastapi import FastAPI, Request
|
|
2 |
from pydantic import BaseModel
|
3 |
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
|
4 |
import torch
|
|
|
|
|
5 |
|
6 |
app = FastAPI()
|
7 |
|
8 |
-
# Load model
|
9 |
model_name = "VietAI/vit5-base-vietnews-summarization"
|
10 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
11 |
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
|
12 |
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
13 |
-
model.to(device)
|
14 |
|
15 |
-
|
16 |
-
class SummaryRequest(BaseModel):
|
17 |
text: str
|
18 |
|
19 |
@app.get("/")
|
20 |
def read_root():
|
21 |
-
return {"message": "
|
22 |
|
23 |
@app.post("/summarize")
|
24 |
-
def summarize(request:
|
25 |
-
|
26 |
-
|
27 |
-
return {"summary": ""}
|
28 |
|
29 |
-
|
30 |
-
|
|
|
|
|
|
|
|
|
31 |
input_ids = encoding["input_ids"].to(device)
|
32 |
attention_mask = encoding["attention_mask"].to(device)
|
33 |
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
|
|
|
|
|
|
41 |
|
42 |
summary = tokenizer.decode(outputs[0], skip_special_tokens=True, clean_up_tokenization_spaces=True)
|
|
|
|
|
|
|
|
|
43 |
return {"summary": summary}
|
|
|
2 |
from pydantic import BaseModel
|
3 |
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
|
4 |
import torch
|
5 |
+
from datetime import datetime
|
6 |
+
import time
|
7 |
|
8 |
app = FastAPI()
|
9 |
|
10 |
+
# Load model and tokenizer
|
11 |
model_name = "VietAI/vit5-base-vietnews-summarization"
|
12 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
13 |
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
|
14 |
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
15 |
+
model = model.to(device)
|
16 |
|
17 |
+
class TextInput(BaseModel):
|
|
|
18 |
text: str
|
19 |
|
20 |
@app.get("/")
|
21 |
def read_root():
|
22 |
+
return {"message": "Summarization API is running"}
|
23 |
|
24 |
@app.post("/summarize")
|
25 |
+
async def summarize(input_text: TextInput, request: Request):
|
26 |
+
start_time = time.time()
|
27 |
+
print(f"[{datetime.now()}] 🔵 Received request from {request.client.host}")
|
|
|
28 |
|
29 |
+
text = input_text.text.strip()
|
30 |
+
prefix = "vietnews: "
|
31 |
+
input_text_prefixed = prefix + text + " </s>"
|
32 |
+
|
33 |
+
# Tokenize
|
34 |
+
encoding = tokenizer(input_text_prefixed, return_tensors="pt", truncation=True, max_length=512)
|
35 |
input_ids = encoding["input_ids"].to(device)
|
36 |
attention_mask = encoding["attention_mask"].to(device)
|
37 |
|
38 |
+
# Generate summary with optimized settings
|
39 |
+
with torch.inference_mode():
|
40 |
+
outputs = model.generate(
|
41 |
+
input_ids=input_ids,
|
42 |
+
attention_mask=attention_mask,
|
43 |
+
max_length=96, # giảm độ dài để xử lý nhanh hơn
|
44 |
+
num_beams=1, # dùng greedy decoding
|
45 |
+
no_repeat_ngram_size=2,
|
46 |
+
early_stopping=True
|
47 |
+
)
|
48 |
|
49 |
summary = tokenizer.decode(outputs[0], skip_special_tokens=True, clean_up_tokenization_spaces=True)
|
50 |
+
|
51 |
+
end_time = time.time()
|
52 |
+
print(f"[{datetime.now()}] ✅ Response sent — total time: {end_time - start_time:.2f}s")
|
53 |
+
|
54 |
return {"summary": summary}
|