Add chatbpot file
Browse files
app.py
ADDED
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import torch
|
2 |
+
from transformers import AutoTokenizer
|
3 |
+
from auto_gptq import AutoGPTQForCausalLM
|
4 |
+
import gradio as gr
|
5 |
+
|
6 |
+
checkpoint = "cortecs/Meta-Llama-3-8B-Instruct-GPTQ-8b"
|
7 |
+
|
8 |
+
# Load tokenizer
|
9 |
+
tokenizer = AutoTokenizer.from_pretrained(checkpoint, use_fast=True)
|
10 |
+
|
11 |
+
# Load GPTQ model
|
12 |
+
|
13 |
+
model = AutoGPTQForCausalLM.from_quantized(
|
14 |
+
checkpoint,
|
15 |
+
device="cuda:0" if torch.cuda.is_available() else "cpu",
|
16 |
+
torch_dtype=torch.float16,
|
17 |
+
)
|
18 |
+
|
19 |
+
# Function to format prompt + generate response
|
20 |
+
def predict(message, history):
|
21 |
+
prompt = f"<s>[INST] {message.strip()} [/INST]"
|
22 |
+
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
|
23 |
+
|
24 |
+
outputs = model.generate(
|
25 |
+
**inputs,
|
26 |
+
do_sample=True,
|
27 |
+
temperature=0.6,
|
28 |
+
top_p=0.9,
|
29 |
+
max_new_tokens=256,
|
30 |
+
eos_token_id=tokenizer.eos_token_id
|
31 |
+
)
|
32 |
+
|
33 |
+
decoded = tokenizer.decode(outputs[0], skip_special_tokens=True)
|
34 |
+
response = decoded.split("[/INST]")[-1].strip()
|
35 |
+
return response
|
36 |
+
|
37 |
+
# Launch Gradio chatbot
|
38 |
+
gr.ChatInterface(predict, title=" LLaMA 3 Chatbot").launch(debug=True)
|