Spaces:
Runtime error
Runtime error
LinkangZhan
commited on
Commit
•
b00538d
1
Parent(s):
d04c3ba
no function
Browse files
app.py
CHANGED
@@ -5,23 +5,48 @@ from threading import Thread
|
|
5 |
import gradio as gr
|
6 |
import torch
|
7 |
|
8 |
-
config = PeftConfig.from_pretrained("Junity/Genshin-World-Model", trust_remote_code=True)
|
9 |
-
model = AutoModelForCausalLM.from_pretrained("baichuan-inc/Baichuan-13B-Base", torch_dtype=torch.float32, trust_remote_code=True)
|
10 |
-
model = PeftModel.from_pretrained(model, r"Junity/Genshin-World-Model", torch_dtype=torch.float32, trust_remote_code=True)
|
11 |
-
tokenizer = AutoTokenizer.from_pretrained("baichuan-inc/Baichuan-13B-Base", trust_remote_code=True)
|
12 |
-
history = []
|
13 |
-
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
14 |
-
if device == "cuda":
|
15 |
-
model.cuda()
|
16 |
-
model = model.half()
|
17 |
|
18 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
if textbox != '':
|
20 |
-
textbox = textbox
|
|
|
|
|
|
|
|
|
|
|
21 |
yield ["", textbox]
|
22 |
else:
|
23 |
-
textbox = textbox
|
|
|
|
|
|
|
|
|
|
|
24 |
yield ["", textbox]
|
|
|
|
|
25 |
input_ids = tokenizer.encode(textbox)[-4096:]
|
26 |
input_ids = torch.LongTensor([input_ids]).to(device)
|
27 |
generation_config = model.generation_config
|
@@ -48,21 +73,32 @@ def respond(role_name, msg, textbox):
|
|
48 |
textbox += new_text
|
49 |
yield ["", textbox]
|
50 |
|
|
|
51 |
with gr.Blocks() as demo:
|
52 |
gr.Markdown(
|
53 |
"""
|
54 |
## Genshin-World-Model
|
55 |
- 模型地址 [https://huggingface.co/Junity/Genshin-World-Model](https://huggingface.co/Junity/Genshin-World-Model)
|
56 |
- 此模型不支持要求对方回答什么,只支持续写。
|
|
|
57 |
"""
|
58 |
)
|
59 |
-
with gr.Tab("
|
60 |
-
role_name = gr.Textbox(label="你将扮演的角色")
|
61 |
-
|
|
|
|
|
|
|
|
|
62 |
with gr.Row():
|
63 |
-
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
|
68 |
-
|
|
|
|
|
|
|
|
|
|
|
|
5 |
import gradio as gr
|
6 |
import torch
|
7 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
|
9 |
+
# lora_folder = ''
|
10 |
+
# model_folder = ''
|
11 |
+
#
|
12 |
+
# config = PeftConfig.from_pretrained(("Junity/Genshin-World-Model" if lora_folder == ''
|
13 |
+
# else lora_folder),
|
14 |
+
# trust_remote_code=True)
|
15 |
+
# model = AutoModelForCausalLM.from_pretrained(("baichuan-inc/Baichuan-13B-Base" if model_folder == ''
|
16 |
+
# else model_folder), torch_dtype=torch.float32, trust_remote_code=True)
|
17 |
+
# model = PeftModel.from_pretrained(model,
|
18 |
+
# ("Junity/Genshin-World-Model" if lora_folder == ''
|
19 |
+
# else lora_folder)
|
20 |
+
# , torch_dtype=torch.float32, trust_remote_code=True)
|
21 |
+
# tokenizer = AutoTokenizer.from_pretrained(("baichuan-inc/Baichuan-13B-Base" if model_folder == ''
|
22 |
+
# else model_folder),
|
23 |
+
# trust_remote_code=True)
|
24 |
+
# history = []
|
25 |
+
# device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
26 |
+
# if device == "cuda":
|
27 |
+
# model.cuda()
|
28 |
+
# model = model.half()
|
29 |
+
|
30 |
+
|
31 |
+
def respond(role_name, character_name, msg, textbox, temp, rep, max_len, top_p, top_k):
|
32 |
if textbox != '':
|
33 |
+
textbox = (textbox
|
34 |
+
+ "\n"
|
35 |
+
+ role_name
|
36 |
+
+ (":" if role_name != '' else '')
|
37 |
+
+ msg
|
38 |
+
+ ('。\n' if msg[-1] not in ['。', '!', '?'] else ''))
|
39 |
yield ["", textbox]
|
40 |
else:
|
41 |
+
textbox = (textbox
|
42 |
+
+ role_name
|
43 |
+
+ (":" if role_name != '' else '')
|
44 |
+
+ msg
|
45 |
+
+ ('。' if msg[-1] not in ['。', '!', '?', ')', '}', ':', ':', '('] else '')
|
46 |
+
+ ('\n' if msg[-1] in ['。', '!', '?', ')', '}'] else ''))
|
47 |
yield ["", textbox]
|
48 |
+
if character_name != '':
|
49 |
+
textbox += ('\n' if textbox[-1] != '\n' else '') + character_name + ':'
|
50 |
input_ids = tokenizer.encode(textbox)[-4096:]
|
51 |
input_ids = torch.LongTensor([input_ids]).to(device)
|
52 |
generation_config = model.generation_config
|
|
|
73 |
textbox += new_text
|
74 |
yield ["", textbox]
|
75 |
|
76 |
+
|
77 |
with gr.Blocks() as demo:
|
78 |
gr.Markdown(
|
79 |
"""
|
80 |
## Genshin-World-Model
|
81 |
- 模型地址 [https://huggingface.co/Junity/Genshin-World-Model](https://huggingface.co/Junity/Genshin-World-Model)
|
82 |
- 此模型不支持要求对方回答什么,只支持续写。
|
83 |
+
- 目前运行不了,因为没有钱租卡。
|
84 |
"""
|
85 |
)
|
86 |
+
with gr.Tab("创作") as chat:
|
87 |
+
role_name = gr.Textbox(label="你将扮演的角色(可留空)")
|
88 |
+
character_name = gr.Textbox(label="对方的角色(可留空)")
|
89 |
+
msg = gr.Textbox(label="你说的话")
|
90 |
+
with gr.Row():
|
91 |
+
clear = gr.ClearButton()
|
92 |
+
sub = gr.Button("Submit", variant="primary")
|
93 |
with gr.Row():
|
94 |
+
temp = gr.Slider(minimum=0, maximum=2.0, step=0.1, value=1.5, label="温度(调大则更随机)", interactive=True)
|
95 |
+
rep = gr.Slider(minimum=0, maximum=2.0, step=0.1, value=1.0, label="对重复生成的惩罚", interactive=True)
|
96 |
+
max_len = gr.Slider(minimum=4, maximum=512, step=4, value=256, label="对方回答的最大长度", interactive=True)
|
97 |
+
top_p = gr.Slider(minimum=0, maximum=1.0, step=0.1, value=0.7, label="Top-p(调大则更随机)", interactive=True)
|
98 |
+
top_k = gr.Slider(minimum=0, maximum=100, step=1, value=50, label="Top-k(调大则更随机)", interactive=True)
|
99 |
+
textbox = gr.Textbox(interactive=True, label="全部文本(可修改)")
|
100 |
+
clear.add([msg, role_name, textbox])
|
101 |
+
sub.click(fn=respond,
|
102 |
+
inputs=[role_name, character_name, msg, textbox, temp, rep, max_len, top_p, top_k],
|
103 |
+
outputs=[msg, textbox])
|
104 |
+
demo.queue().launch(server_port=6006, share=True)
|