Spaces:
Runtime error
Runtime error
Commit
·
b96ad8e
1
Parent(s):
70226f5
update app.py
Browse files
app.py
CHANGED
@@ -23,65 +23,6 @@ download_file(ggml_model_path, filename)
|
|
23 |
llm = Llama(model_path=filename, n_ctx=512, n_batch=126)
|
24 |
|
25 |
|
26 |
-
def generate_text(prompt="Who is the CEO of Apple?"):
|
27 |
-
output = llm(
|
28 |
-
prompt,
|
29 |
-
max_tokens=256,
|
30 |
-
temperature=0.1,
|
31 |
-
top_p=0.5,
|
32 |
-
echo=False,
|
33 |
-
stop=["#"],
|
34 |
-
)
|
35 |
-
output_text = output["choices"][0]["text"].strip()
|
36 |
-
|
37 |
-
# Remove Prompt Echo from Generated Text
|
38 |
-
cleaned_output_text = output_text.replace(prompt, "")
|
39 |
-
return cleaned_output_text
|
40 |
-
|
41 |
-
|
42 |
-
description = "Vicuna-7B"
|
43 |
-
|
44 |
-
examples = [
|
45 |
-
["What is the capital of France?", "The capital of France is Paris."],
|
46 |
-
[
|
47 |
-
"Who wrote the novel 'Pride and Prejudice'?",
|
48 |
-
"The novel 'Pride and Prejudice' was written by Jane Austen.",
|
49 |
-
],
|
50 |
-
["What is the square root of 64?", "The square root of 64 is 8."],
|
51 |
-
]
|
52 |
-
|
53 |
-
gradio_interface = gr.Interface(
|
54 |
-
fn=generate_text,
|
55 |
-
inputs="text",
|
56 |
-
outputs="text",
|
57 |
-
examples=examples,
|
58 |
-
title="Vicuna-7B",
|
59 |
-
)
|
60 |
-
gradio_interface.launch()import os
|
61 |
-
import urllib.request
|
62 |
-
import gradio as gr
|
63 |
-
from llama_cpp import Llama
|
64 |
-
|
65 |
-
|
66 |
-
def download_file(file_link, filename):
|
67 |
-
# Checks if the file already exists before downloading
|
68 |
-
if not os.path.isfile(filename):
|
69 |
-
urllib.request.urlretrieve(file_link, filename)
|
70 |
-
print("File downloaded successfully.")
|
71 |
-
else:
|
72 |
-
print("File already exists.")
|
73 |
-
|
74 |
-
|
75 |
-
# Dowloading GGML model from HuggingFace
|
76 |
-
ggml_model_path = "https://huggingface.co/CRD716/ggml-vicuna-1.1-quantized/resolve/main/ggml-vicuna-7b-1.1-q4_1.bin"
|
77 |
-
filename = "ggml-vicuna-7b-1.1-q4_1.bin"
|
78 |
-
|
79 |
-
download_file(ggml_model_path, filename)
|
80 |
-
|
81 |
-
|
82 |
-
llm = Llama(model_path=filename, n_ctx=512, n_batch=126)
|
83 |
-
|
84 |
-
|
85 |
def generate_text(prompt="Who is the CEO of Apple?"):
|
86 |
output = llm(
|
87 |
prompt,
|
|
|
23 |
llm = Llama(model_path=filename, n_ctx=512, n_batch=126)
|
24 |
|
25 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
26 |
def generate_text(prompt="Who is the CEO of Apple?"):
|
27 |
output = llm(
|
28 |
prompt,
|