Update app.py
Browse files
app.py
CHANGED
@@ -24,103 +24,94 @@ hf_token = os.getenv('HF_API_TOKEN')
|
|
24 |
if not hf_token:
|
25 |
raise ValueError("Hugging Face API token is not set. Please set the HF_API_TOKEN environment variable.")
|
26 |
|
27 |
-
|
|
|
|
|
28 |
model_gpt2 = GPT2LMHeadModel.from_pretrained('gpt2')
|
|
|
|
|
29 |
text_generator = pipeline("text-generation", model=model_gpt2, tokenizer=tokenizer_gpt2)
|
30 |
|
|
|
31 |
prompt_template = """\
|
32 |
-
You are an
|
33 |
|
34 |
-
|
35 |
|
36 |
Description: {description}
|
37 |
|
38 |
-
|
39 |
-
|
40 |
-
Please provide the data in CSV format.
|
41 |
-
|
42 |
-
Example Description:
|
43 |
-
Generate a dataset for predicting house prices with columns: 'Size', 'Location', 'Number of Bedrooms', 'Price'
|
44 |
|
45 |
-
|
46 |
-
Size,Location,Number of Bedrooms,Price
|
47 |
-
1200,Suburban,3,250000
|
48 |
-
900,Urban,2,200000
|
49 |
-
1500,Rural,4,300000
|
50 |
-
...
|
51 |
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
Output: """
|
57 |
|
58 |
-
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
-
|
63 |
-
|
64 |
-
def format_prompt(description, columns):
|
65 |
-
processed_description = preprocess_user_prompt(description)
|
66 |
-
prompt = prompt_template.format(description=processed_description, columns=",".join(columns))
|
67 |
-
return prompt
|
68 |
|
69 |
-
API_URL = "https://api-inference.huggingface.co/models/mistralai/Mixtral-8x7B-Instruct-v0.1"
|
70 |
|
|
|
71 |
generation_params = {
|
72 |
"top_p": 0.90,
|
73 |
"temperature": 0.8,
|
74 |
-
"max_new_tokens":
|
75 |
"return_full_text": False,
|
76 |
"use_cache": False
|
77 |
}
|
78 |
|
|
|
|
|
|
|
|
|
79 |
def generate_synthetic_data(description, columns):
|
80 |
formatted_prompt = format_prompt(description, columns)
|
81 |
payload = {"inputs": formatted_prompt, "parameters": generation_params}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
82 |
try:
|
83 |
-
|
84 |
-
response.raise_for_status()
|
85 |
-
data = response.json()
|
86 |
-
if 'generated_text' in data[0]:
|
87 |
-
return data[0]['generated_text']
|
88 |
-
else:
|
89 |
-
raise ValueError("Invalid response format from Hugging Face API.")
|
90 |
-
except (requests.RequestException, ValueError) as e:
|
91 |
-
print(f"Error during API request or response processing: {e}")
|
92 |
-
return ""
|
93 |
-
|
94 |
-
def process_generated_data(csv_data, expected_columns):
|
95 |
-
try:
|
96 |
-
# Replace inconsistent line endings
|
97 |
cleaned_data = csv_data.replace('\r\n', '\n').replace('\r', '\n')
|
98 |
-
|
99 |
-
# Check for common CSV formatting issues and apply corrections
|
100 |
-
cleaned_data = cleaned_data.strip().replace('|', ',').replace(' ', ' ').replace(' ,', ',')
|
101 |
-
|
102 |
-
# Load the cleaned data into a DataFrame
|
103 |
data = StringIO(cleaned_data)
|
104 |
-
|
105 |
-
|
|
|
|
|
106 |
return df
|
107 |
-
|
108 |
except pd.errors.ParserError as e:
|
109 |
print(f"Failed to parse CSV data: {e}")
|
110 |
return None
|
111 |
|
112 |
def generate_large_synthetic_data(description, columns, num_rows=1000, rows_per_generation=100):
|
113 |
data_frames = []
|
114 |
-
|
115 |
for _ in tqdm(range(num_rows // rows_per_generation), desc="Generating Data"):
|
116 |
generated_data = generate_synthetic_data(description, columns)
|
117 |
-
df_synthetic = process_generated_data(generated_data, columns)
|
118 |
|
119 |
-
if
|
120 |
-
|
121 |
-
|
122 |
-
|
123 |
-
|
|
|
|
|
|
|
|
|
124 |
if data_frames:
|
125 |
return pd.concat(data_frames, ignore_index=True)
|
126 |
else:
|
|
|
24 |
if not hf_token:
|
25 |
raise ValueError("Hugging Face API token is not set. Please set the HF_API_TOKEN environment variable.")
|
26 |
|
27 |
+
|
28 |
+
# Load GPT-2 model and tokenizer
|
29 |
+
tokenizer_gpt2 = AutoTokenizer.from_pretrained('gpt2')
|
30 |
model_gpt2 = GPT2LMHeadModel.from_pretrained('gpt2')
|
31 |
+
|
32 |
+
# Create a pipeline for text generation using GPT-2
|
33 |
text_generator = pipeline("text-generation", model=model_gpt2, tokenizer=tokenizer_gpt2)
|
34 |
|
35 |
+
# Define prompt template for generating the dataset
|
36 |
prompt_template = """\
|
37 |
+
You are an AI specialized in generating synthetic tabular data specifically for machine learning purposes.
|
38 |
|
39 |
+
Task: Generate a synthetic dataset based on the provided description and column names.
|
40 |
|
41 |
Description: {description}
|
42 |
|
43 |
+
Columns: {columns}
|
|
|
|
|
|
|
|
|
|
|
44 |
|
45 |
+
Instructions:
|
|
|
|
|
|
|
|
|
|
|
46 |
|
47 |
+
Output only the tabular data in valid CSV format.
|
48 |
+
Include the header row followed by the data rows.
|
49 |
+
Do not generate any additional text, explanations, comments, or code.
|
50 |
+
Ensure that the values for each column are contextually appropriate.
|
|
|
51 |
|
52 |
+
Format Example (do not include this line or the following example in your output):
|
53 |
+
Column1,Column2,Column3
|
54 |
+
Value1,Value2,Value3
|
55 |
+
Value4,Value5,Value6
|
56 |
+
"""
|
|
|
|
|
|
|
|
|
|
|
57 |
|
|
|
58 |
|
59 |
+
# Define generation parameters
|
60 |
generation_params = {
|
61 |
"top_p": 0.90,
|
62 |
"temperature": 0.8,
|
63 |
+
"max_new_tokens": 1024,
|
64 |
"return_full_text": False,
|
65 |
"use_cache": False
|
66 |
}
|
67 |
|
68 |
+
def format_prompt(description, columns):
|
69 |
+
prompt = prompt_template.format(description=description, columns=",".join(columns))
|
70 |
+
return prompt
|
71 |
+
|
72 |
def generate_synthetic_data(description, columns):
|
73 |
formatted_prompt = format_prompt(description, columns)
|
74 |
payload = {"inputs": formatted_prompt, "parameters": generation_params}
|
75 |
+
|
76 |
+
# Call Mixtral model to generate data
|
77 |
+
response = requests.post("https://api-inference.huggingface.co/models/mistralai/Mixtral-8x7B-Instruct-v0.1",
|
78 |
+
headers={"Authorization": f"Bearer {token}"}, json=payload)
|
79 |
+
|
80 |
+
if response.status_code == 200:
|
81 |
+
return response.json()[0]["generated_text"]
|
82 |
+
else:
|
83 |
+
print(f"Error generating data: {response.status_code}, {response.text}")
|
84 |
+
return None
|
85 |
+
|
86 |
+
def process_generated_data(csv_data):
|
87 |
try:
|
88 |
+
# Ensure the data is cleaned and correctly formatted
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
89 |
cleaned_data = csv_data.replace('\r\n', '\n').replace('\r', '\n')
|
|
|
|
|
|
|
|
|
|
|
90 |
data = StringIO(cleaned_data)
|
91 |
+
|
92 |
+
# Read the CSV data with specific parameters to handle irregularities
|
93 |
+
df = pd.read_csv(data)
|
94 |
+
|
95 |
return df
|
|
|
96 |
except pd.errors.ParserError as e:
|
97 |
print(f"Failed to parse CSV data: {e}")
|
98 |
return None
|
99 |
|
100 |
def generate_large_synthetic_data(description, columns, num_rows=1000, rows_per_generation=100):
|
101 |
data_frames = []
|
102 |
+
|
103 |
for _ in tqdm(range(num_rows // rows_per_generation), desc="Generating Data"):
|
104 |
generated_data = generate_synthetic_data(description, columns)
|
|
|
105 |
|
106 |
+
if generated_data:
|
107 |
+
|
108 |
+
df_synthetic = process_generated_data(generated_data)
|
109 |
+
|
110 |
+
if df_synthetic is not None and not df_synthetic.empty:
|
111 |
+
data_frames.append(df_synthetic)
|
112 |
+
else:
|
113 |
+
print("Skipping invalid generation.")
|
114 |
+
|
115 |
if data_frames:
|
116 |
return pd.concat(data_frames, ignore_index=True)
|
117 |
else:
|