Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -7,13 +7,6 @@ import json
|
|
7 |
import subprocess
|
8 |
import sys
|
9 |
from llama_cpp import Llama,llama_model_decoder_start_token
|
10 |
-
from llama_cpp_agent import LlamaCppAgent
|
11 |
-
from llama_cpp_agent import MessagesFormatterType
|
12 |
-
from llama_cpp_agent.providers import LlamaCppPythonProvider
|
13 |
-
from llama_cpp_agent.chat_history import BasicChatHistory
|
14 |
-
from llama_cpp_agent.chat_history.messages import Roles
|
15 |
-
from llama_cpp_agent.chat_history.messages import Roles
|
16 |
-
from llama_cpp_agent.messages_formatter import MessagesFormatter, PromptMarkers
|
17 |
import gradio as gr
|
18 |
from huggingface_hub import hf_hub_download
|
19 |
from typing import List, Tuple
|
@@ -77,6 +70,8 @@ def respond(
|
|
77 |
Returns:
|
78 |
str: The response to the message.
|
79 |
"""
|
|
|
|
|
80 |
try:
|
81 |
global llama
|
82 |
if llama == None:
|
@@ -85,7 +80,7 @@ def respond(
|
|
85 |
n_batch=16,
|
86 |
n_ctx=512,
|
87 |
n_threads=2,
|
88 |
-
n_threads_batch=2,
|
89 |
|
90 |
tokens = llama.tokenize(f"<2ja>{message}".encode("utf-8"))
|
91 |
llama.encode(tokens)
|
|
|
7 |
import subprocess
|
8 |
import sys
|
9 |
from llama_cpp import Llama,llama_model_decoder_start_token
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
10 |
import gradio as gr
|
11 |
from huggingface_hub import hf_hub_download
|
12 |
from typing import List, Tuple
|
|
|
70 |
Returns:
|
71 |
str: The response to the message.
|
72 |
"""
|
73 |
+
if model==None:
|
74 |
+
return
|
75 |
try:
|
76 |
global llama
|
77 |
if llama == None:
|
|
|
80 |
n_batch=16,
|
81 |
n_ctx=512,
|
82 |
n_threads=2,
|
83 |
+
n_threads_batch=2,verbose=False)
|
84 |
|
85 |
tokens = llama.tokenize(f"<2ja>{message}".encode("utf-8"))
|
86 |
llama.encode(tokens)
|