cistine commited on
Commit
faf20f4
1 Parent(s): 2ae29ec

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +58 -1
app.py CHANGED
@@ -1,3 +1,60 @@
 
1
  import gradio as gr
 
 
 
 
2
 
3
- gr.load("models/bigcode/starcoder2-3b").launch()
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import spaces
2
  import gradio as gr
3
+ import torch
4
+ import transformers
5
+ from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
6
+ import os
7
 
8
+ title = """# Welcome to 🌟Tonic's✨StarCoder
9
+ ✨StarCoder StarCoder2-15B model is a 15B parameter model trained on 600+ programming languages from The Stack v2, with opt-out requests excluded. The model uses Grouped Query Attention, a context window of 16,384 tokens with a sliding window attention of 4,096 tokens, and was trained using the Fill-in-the-Middle objective on 4+ trillion tokens. The model was trained with NVIDIA NeMo™ Framework using the NVIDIA Eos Supercomputer built with NVIDIA DGX H100 systems. You can build with this endpoint using✨StarCoder available here : [bigcode/starcoder2-15b](https://huggingface.co/bigcode/starcoder2-15b). You can also use ✨StarCoder by cloning this space. Simply click here: <a style="display:inline-block" href="https://huggingface.co/spaces/Tonic/starcoder2?duplicate=true"><img src="https://img.shields.io/badge/-Duplicate%20Space-blue?labelColor=white&style=flat&logo=data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABAAAAAQCAYAAAAf8/9hAAAAAXNSR0IArs4c6QAAAP5JREFUOE+lk7FqAkEURY+ltunEgFXS2sZGIbXfEPdLlnxJyDdYB62sbbUKpLbVNhyYFzbrrA74YJlh9r079973psed0cvUD4A+4HoCjsA85X0Dfn/RBLBgBDxnQPfAEJgBY+A9gALA4tcbamSzS4xq4FOQAJgCDwV2CPKV8tZAJcAjMMkUe1vX+U+SMhfAJEHasQIWmXNN3abzDwHUrgcRGmYcgKe0bxrblHEB4E/pndMazNpSZGcsZdBlYJcEL9Afo75molJyM2FxmPgmgPqlWNLGfwZGG6UiyEvLzHYDmoPkDDiNm9JR9uboiONcBXrpY1qmgs21x1QwyZcpvxt9NS09PlsPAAAAAElFTkSuQmCC&logoWidth=14" alt="Duplicate Space"></a></h3>
10
+ Join us : 🌟TeamTonic🌟 is always making cool demos! Join our active builder's 🛠️community 👻 [![Join us on Discord](https://img.shields.io/discord/1109943800132010065?label=Discord&logo=discord&style=flat-square)](https://discord.gg/GWpVpekp) On 🤗Huggingface:[MultiTransformer](https://huggingface.co/MultiTransformer) Math 🔍 [introspector](https://huggingface.co/introspector) On 🌐Github: [Tonic-AI](https://github.com/tonic-ai) & contribute to🌟 [SciTonic](https://github.com/Tonic-AI/scitonic)🤗Big thanks to Yuvi Sharma and all the folks at huggingface for the community grant 🤗
11
+ """
12
+
13
+ model_path = "bigcode/starcoder2-3b"
14
+
15
+ hf_token = os.getenv("HF_TOKEN")
16
+ if not hf_token:
17
+ raise ValueError("Hugging Face token not found. Please set the HF_TOKEN environment variable.")
18
+
19
+ tokenizer = AutoTokenizer.from_pretrained(model_path)
20
+ quantization_config = BitsAndBytesConfig(load_in_8bit=True)
21
+ model = AutoModelForCausalLM.from_pretrained( model_path, quantization_config=quantization_config)
22
+
23
+ @spaces.GPU
24
+ def generate_text(prompt, temperature=0.9, max_length=1200):
25
+ # Encode the inputs
26
+ inputs = tokenizer.encode(prompt, return_tensors="pt")
27
+ attention_mask = torch.ones(inputs.shape, dtype=torch.long)
28
+ inputs = inputs.to("cuda")
29
+ attention_mask = attention_mask.to("cuda")
30
+ outputs = model.generate(
31
+ inputs,
32
+ attention_mask=attention_mask,
33
+ max_length=max_length,
34
+ top_p=0.9,
35
+ temperature=temperature,
36
+ do_sample=True,
37
+ pad_token_id=tokenizer.eos_token_id
38
+ )
39
+ return tokenizer.decode(outputs[0])
40
+
41
+ def gradio_app():
42
+ with gr.Blocks() as demo:
43
+ gr.Markdown(title)
44
+ prompt = gr.Code(label="Enter your code prompt", value="def print_hello_world():")
45
+ with gr.Row():
46
+ temperature = gr.Slider(minimum=0.1, maximum=1.0, step=0.1, value=0.5, label="Temperature")
47
+ max_length = gr.Slider(minimum=100, maximum=1024, step=10, value=200, label="Generate Length")
48
+ generate_btn = gr.Button("Try✨StarCoder")
49
+ output = gr.Code(label="✨StarCoder:", lines=40)
50
+
51
+ generate_btn.click(
52
+ fn=generate_text,
53
+ inputs=[prompt, temperature, max_length],
54
+ outputs=output
55
+ )
56
+
57
+ demo.launch()
58
+
59
+ if __name__ == "__main__":
60
+ gradio_app()