Update app.py
Browse files
app.py
CHANGED
@@ -16,12 +16,6 @@ headers = {
|
|
16 |
# Base system message
|
17 |
BASE_SYSTEM_MESSAGE = "I carefully provide accurate, factual, thoughtful, nuanced answers and am brilliant at reasoning."
|
18 |
|
19 |
-
def clear_chat():
|
20 |
-
"""Clears the chat history and message state."""
|
21 |
-
print("Clearing chat...")
|
22 |
-
chat_history_state.value = []
|
23 |
-
chatbot.textbox.value = ""
|
24 |
-
|
25 |
def call_nvidia_api(history, system_message, max_tokens, temperature, top_p):
|
26 |
"""Calls the NVIDIA API to generate a response."""
|
27 |
messages = [{"role": "system", "content": system_message}]
|
@@ -34,7 +28,7 @@ def call_nvidia_api(history, system_message, max_tokens, temperature, top_p):
|
|
34 |
"max_tokens": max_tokens,
|
35 |
"stream": False
|
36 |
}
|
37 |
-
print(f"Payload enviado: {payload}")
|
38 |
session = requests.Session()
|
39 |
response = session.post(INVOKE_URL, headers=headers, json=payload)
|
40 |
while response.status_code == 202:
|
@@ -43,7 +37,7 @@ def call_nvidia_api(history, system_message, max_tokens, temperature, top_p):
|
|
43 |
response = session.get(fetch_url, headers=headers)
|
44 |
response.raise_for_status()
|
45 |
response_body = response.json()
|
46 |
-
print(f"Payload recebido: {response_body}")
|
47 |
if response_body.get("choices"):
|
48 |
assistant_message = response_body["choices"][0]["message"]["content"]
|
49 |
return assistant_message
|
@@ -70,6 +64,7 @@ system_msg = gr.Textbox(BASE_SYSTEM_MESSAGE, label="System Message", placeholder
|
|
70 |
max_tokens = gr.Slider(20, 1024, label="Max Tokens", step=20, value=1024)
|
71 |
temperature = gr.Slider(0.0, 1.0, label="Temperature", step=0.1, value=0.2)
|
72 |
top_p = gr.Slider(0.0, 1.0, label="Top P", step=0.05, value=0.7)
|
|
|
73 |
with gr.Blocks() as demo:
|
74 |
chat_history_state = gr.State([])
|
75 |
chatbot = gr.ChatInterface(
|
|
|
16 |
# Base system message
|
17 |
BASE_SYSTEM_MESSAGE = "I carefully provide accurate, factual, thoughtful, nuanced answers and am brilliant at reasoning."
|
18 |
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
def call_nvidia_api(history, system_message, max_tokens, temperature, top_p):
|
20 |
"""Calls the NVIDIA API to generate a response."""
|
21 |
messages = [{"role": "system", "content": system_message}]
|
|
|
28 |
"max_tokens": max_tokens,
|
29 |
"stream": False
|
30 |
}
|
31 |
+
print(f"Payload enviado: {json.dumps(payload, indent=2)}")
|
32 |
session = requests.Session()
|
33 |
response = session.post(INVOKE_URL, headers=headers, json=payload)
|
34 |
while response.status_code == 202:
|
|
|
37 |
response = session.get(fetch_url, headers=headers)
|
38 |
response.raise_for_status()
|
39 |
response_body = response.json()
|
40 |
+
print(f"Payload recebido: {json.dumps(response_body, indent=2)}")
|
41 |
if response_body.get("choices"):
|
42 |
assistant_message = response_body["choices"][0]["message"]["content"]
|
43 |
return assistant_message
|
|
|
64 |
max_tokens = gr.Slider(20, 1024, label="Max Tokens", step=20, value=1024)
|
65 |
temperature = gr.Slider(0.0, 1.0, label="Temperature", step=0.1, value=0.2)
|
66 |
top_p = gr.Slider(0.0, 1.0, label="Top P", step=0.05, value=0.7)
|
67 |
+
|
68 |
with gr.Blocks() as demo:
|
69 |
chat_history_state = gr.State([])
|
70 |
chatbot = gr.ChatInterface(
|