Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -7,26 +7,13 @@ from huggingface_hub import InferenceClient
|
|
7 |
"""
|
8 |
For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
|
9 |
"""
|
10 |
-
|
11 |
|
12 |
-
client = InferenceClient(
|
13 |
-
|
14 |
-
|
15 |
-
)
|
16 |
-
|
17 |
-
# completion = client.chat.completions.create(
|
18 |
-
# model="deepseek-ai/DeepSeek-V3-0324",
|
19 |
-
# messages=[
|
20 |
-
# {
|
21 |
-
# "role": "user",
|
22 |
-
# "content": "What is the capital of France?"
|
23 |
-
# }
|
24 |
-
# ],
|
25 |
-
# max_tokens=500,
|
26 |
# )
|
27 |
|
28 |
-
# print(completion.choices[0].message)
|
29 |
-
|
30 |
global history
|
31 |
history = []
|
32 |
async def respond(
|
@@ -47,26 +34,26 @@ async def respond(
|
|
47 |
|
48 |
messages.append({"role": "user", "content": message})
|
49 |
|
50 |
-
|
51 |
-
|
52 |
-
#for message in client.chat_completion(
|
53 |
-
# model="deepseek-ai/DeepSeek-V3-0324",
|
54 |
-
# messages=messages,
|
55 |
-
# max_tokens=max_tokens,
|
56 |
-
# stream=True,
|
57 |
-
# temperature=temperature,
|
58 |
-
# top_p=top_p,
|
59 |
-
# ):
|
60 |
-
# token = message.choices[0].delta.content
|
61 |
-
# response += token
|
62 |
|
63 |
-
|
64 |
-
model="
|
65 |
messages=messages,
|
66 |
-
max_tokens=
|
67 |
-
|
68 |
-
|
69 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
70 |
|
71 |
communicate = edge_tts.Communicate(response, voice="mn-MN-YesuiNeural")
|
72 |
with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_file:
|
|
|
7 |
"""
|
8 |
For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
|
9 |
"""
|
10 |
+
client = InferenceClient("google/gemma-3-27b-it", token=os.getenv("TOKEN"))
|
11 |
|
12 |
+
# client = InferenceClient(
|
13 |
+
# provider="fireworks-ai",
|
14 |
+
# api_key=os.getenv("TOKEN"),
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
15 |
# )
|
16 |
|
|
|
|
|
17 |
global history
|
18 |
history = []
|
19 |
async def respond(
|
|
|
34 |
|
35 |
messages.append({"role": "user", "content": message})
|
36 |
|
37 |
+
response = ""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
38 |
|
39 |
+
for message in client.chat_completion(
|
40 |
+
model="google/gemma-3-27b-it",
|
41 |
messages=messages,
|
42 |
+
max_tokens=max_tokens,
|
43 |
+
stream=True,
|
44 |
+
temperature=temperature,
|
45 |
+
top_p=top_p,
|
46 |
+
):
|
47 |
+
token = message.choices[0].delta.content
|
48 |
+
response += token
|
49 |
+
|
50 |
+
# completion = client.chat.completions.create(
|
51 |
+
# model="deepseek-ai/DeepSeek-R1",
|
52 |
+
# messages=messages,
|
53 |
+
# max_tokens=500,
|
54 |
+
# )
|
55 |
+
# response = completion.choices[0].message.content
|
56 |
+
# print(response)
|
57 |
|
58 |
communicate = edge_tts.Communicate(response, voice="mn-MN-YesuiNeural")
|
59 |
with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_file:
|