Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
@@ -10,6 +10,7 @@ model = ""
|
|
10 |
duration = None
|
11 |
token = os.getenv('deepseekv2')
|
12 |
provider = None #'fal-ai' #None #replicate # sambanova
|
|
|
13 |
|
14 |
print(f"Is CUDA available: {torch.cuda.is_available()}")
|
15 |
print(f"CUDA device: {torch.cuda.get_device_name(torch.cuda.current_device())}")
|
@@ -47,6 +48,10 @@ def choose_model(model_name):
|
|
47 |
elif model_name == "Llama2-13b-chat":
|
48 |
model = "meta-llama/Llama-2-13b-chat-hf"
|
49 |
|
|
|
|
|
|
|
|
|
50 |
elif model_name == "Gemma-2-2b":
|
51 |
model = "google/gemma-2-2b-it"
|
52 |
|
@@ -94,23 +99,41 @@ def respond(message, history: list[tuple[str, str]], model, system_message, max_
|
|
94 |
response += token
|
95 |
yield response
|
96 |
|
97 |
-
|
98 |
-
demo = gr.ChatInterface(
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
|
104 |
-
|
105 |
-
|
106 |
-
|
107 |
-
|
108 |
-
|
109 |
-
|
110 |
-
|
111 |
-
|
112 |
-
|
113 |
-
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
114 |
|
115 |
|
116 |
if __name__ == "__main__":
|
|
|
10 |
duration = None
|
11 |
token = os.getenv('deepseekv2')
|
12 |
provider = None #'fal-ai' #None #replicate # sambanova
|
13 |
+
mode = "text-to-text"
|
14 |
|
15 |
print(f"Is CUDA available: {torch.cuda.is_available()}")
|
16 |
print(f"CUDA device: {torch.cuda.get_device_name(torch.cuda.current_device())}")
|
|
|
48 |
elif model_name == "Llama2-13b-chat":
|
49 |
model = "meta-llama/Llama-2-13b-chat-hf"
|
50 |
|
51 |
+
elif model_name == "Llama-3.2-11B-Vision-Instruct":
|
52 |
+
model = "meta-llama/Llama-3.2-11B-Vision-Instruct"
|
53 |
+
mode = "image-to-text"
|
54 |
+
|
55 |
elif model_name == "Gemma-2-2b":
|
56 |
model = "google/gemma-2-2b-it"
|
57 |
|
|
|
99 |
response += token
|
100 |
yield response
|
101 |
|
102 |
+
if mode == "text-to-text":
|
103 |
+
demo = gr.ChatInterface(
|
104 |
+
respond,
|
105 |
+
stop_btn = "Stop generation",
|
106 |
+
multimodal = True,
|
107 |
+
title="Ask me anything",
|
108 |
+
description="Hi there! I am your friendly AI chatbot. Choose from different language models under the Additional Inputs tab below.",
|
109 |
+
examples=[["Explain quantum computing"], ["Explain forex trading"], ["What is the capital of China?"], ["Make a poem about nature"]],
|
110 |
+
additional_inputs=[
|
111 |
+
gr.Dropdown(["DeepSeek-R1-Distill-Qwen-1.5B", "DeepSeek-R1-Distill-Qwen-32B", "Gemma-2-2b", "Gemma-7b", "Llama2-13b-chat", "Llama3-8b-Instruct", "Llama3.1-8b-Instruct", "Microsoft-phi-2", "Mixtral-8x7B-Instruct", "Qwen2.5-Coder-32B-Instruct", "Zephyr-7b-beta"], label="Select Model"),
|
112 |
+
gr.Textbox(value="You are a friendly and helpful Chatbot, be concise and straight to the point, avoid excessive reasoning.", label="System message"),
|
113 |
+
gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
|
114 |
+
gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
|
115 |
+
gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)")
|
116 |
+
|
117 |
+
]
|
118 |
+
)
|
119 |
+
|
120 |
+
elif mode == "image-to-text":
|
121 |
+
demo = gr.ChatInterface(
|
122 |
+
respond,
|
123 |
+
stop_btn = "Stop generation",
|
124 |
+
multimodal = True,
|
125 |
+
title="Ask me anything",
|
126 |
+
description="Hi there! I am your friendly AI chatbot. Choose from different language models under the Additional Inputs tab below.",
|
127 |
+
examples=[["Explain quantum computing"], ["Explain forex trading"], ["What is the capital of China?"], ["Make a poem about nature"]],
|
128 |
+
additional_inputs=[
|
129 |
+
gr.Dropdown(["DeepSeek-R1-Distill-Qwen-1.5B", "DeepSeek-R1-Distill-Qwen-32B", "Gemma-2-2b", "Gemma-7b", "Llama2-13b-chat", "Llama3-8b-Instruct", "Llama3.1-8b-Instruct", "Llama-3.2-11B-Vision-Instruct", "Microsoft-phi-2", "Mixtral-8x7B-Instruct", "Qwen2.5-Coder-32B-Instruct", "Zephyr-7b-beta"], label="Select Model"),
|
130 |
+
gr.Textbox(value="You are a friendly and helpful Chatbot, be concise and straight to the point, avoid excessive reasoning.", label="System message"),
|
131 |
+
gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
|
132 |
+
gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
|
133 |
+
gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)")
|
134 |
+
|
135 |
+
]
|
136 |
+
)
|
137 |
|
138 |
|
139 |
if __name__ == "__main__":
|