Spaces:
Running
Running
File size: 10,186 Bytes
0611c31 2cb9dec fdc226e 0611c31 2cb9dec bc82930 2cb9dec bc82930 e0b1978 2cb9dec bc82930 e0b1978 bc82930 2cb9dec fdc226e 0611c31 fdc226e 0611c31 fdc226e 0611c31 fdc226e 0611c31 fdc226e 0611c31 fdc226e 0611c31 fdc226e 0611c31 fdc226e 0611c31 fdc226e 0611c31 e0b1978 0611c31 2cb9dec fdc226e e0b1978 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 |
# from openai import AsyncOpenAI
# import logging
# from typing import List, Dict, Union
# import pandas as pd
# import asyncio
# from src.api.exceptions import OpenAIError
# # Set up structured logging
# logging.basicConfig(
# level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
# )
# logger = logging.getLogger(__name__)
# class EmbeddingService:
# def __init__(
# self,
# openai_api_key: str,
# model: str = "text-embedding-3-small",
# batch_size: int = 10,
# max_concurrent_requests: int = 10, # Limit to 10 concurrent requests
# ):
# self.client = AsyncOpenAI(api_key=openai_api_key)
# self.model = model
# self.batch_size = batch_size
# self.semaphore = asyncio.Semaphore(max_concurrent_requests) # Rate limiter
# self.total_requests = 0 # Total number of requests to process
# self.completed_requests = 0 # Number of completed requests
# async def get_embedding(self, text: str) -> List[float]:
# """Generate embeddings for the given text using OpenAI."""
# text = text.replace("\n", " ")
# try:
# async with self.semaphore: # Acquire a semaphore slot
# response = await self.client.embeddings.create(
# input=[text], model=self.model
# )
# self.completed_requests += 1 # Increment completed requests
# self._log_progress() # Log progress
# return response.data[0].embedding
# except Exception as e:
# logger.error(f"Failed to generate embedding: {e}")
# raise OpenAIError(f"OpenAI API error: {e}")
# async def create_embeddings(
# self,
# data: Union[pd.DataFrame, List[str]],
# target_column: str = None,
# output_column: str = "embeddings",
# ) -> Union[pd.DataFrame, List[List[float]]]:
# """
# Create embeddings for either a DataFrame or a list of strings.
# Args:
# data: Either a DataFrame or a list of strings.
# target_column: The column in the DataFrame to generate embeddings for (required if data is a DataFrame).
# output_column: The column to store embeddings in the DataFrame (default: "embeddings").
# Returns:
# If data is a DataFrame, returns the DataFrame with the embeddings column.
# If data is a list of strings, returns a list of embeddings.
# """
# if isinstance(data, pd.DataFrame):
# if not target_column:
# raise ValueError("target_column is required when data is a DataFrame.")
# return await self._create_embeddings_for_dataframe(
# data, target_column, output_column
# )
# elif isinstance(data, list):
# return await self._create_embeddings_for_texts(data)
# else:
# raise TypeError(
# "data must be either a pandas DataFrame or a list of strings."
# )
# async def _create_embeddings_for_dataframe(
# self, df: pd.DataFrame, target_column: str, output_column: str
# ) -> pd.DataFrame:
# """Create embeddings for the target column in the DataFrame."""
# logger.info("Generating embeddings for DataFrame...")
# self.total_requests = len(df) # Set total number of requests
# self.completed_requests = 0 # Reset completed requests counter
# batches = [
# df[i : i + self.batch_size] for i in range(0, len(df), self.batch_size)
# ]
# processed_batches = await asyncio.gather(
# *[
# self._process_batch(batch, target_column, output_column)
# for batch in batches
# ]
# )
# return pd.concat(processed_batches)
# async def _create_embeddings_for_texts(self, texts: List[str]) -> List[List[float]]:
# """Create embeddings for a list of strings."""
# logger.info("Generating embeddings for list of texts...")
# self.total_requests = len(texts) # Set total number of requests
# self.completed_requests = 0 # Reset completed requests counter
# batches = [
# texts[i : i + self.batch_size]
# for i in range(0, len(texts), self.batch_size)
# ]
# embeddings = []
# for batch in batches:
# batch_embeddings = await asyncio.gather(
# *[self.get_embedding(text) for text in batch]
# )
# embeddings.extend(batch_embeddings)
# return embeddings
# async def _process_batch(
# self, df_batch: pd.DataFrame, target_column: str, output_column: str
# ) -> pd.DataFrame:
# """Process a batch of rows to generate embeddings."""
# embeddings = await asyncio.gather(
# *[self.get_embedding(row[target_column]) for _, row in df_batch.iterrows()]
# )
# df_batch[output_column] = embeddings
# return df_batch
# def _log_progress(self):
# """Log the progress of embedding generation."""
# progress = (self.completed_requests / self.total_requests) * 100
# logger.info(
# f"Progress: {self.completed_requests}/{self.total_requests} ({progress:.2f}%)"
# )
from openai import AsyncOpenAI
import logging
from typing import List, Dict, Union
from datasets import Dataset
import asyncio
from src.api.exceptions import OpenAIError
# Set up structured logging
logging.basicConfig(
level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
)
logger = logging.getLogger(__name__)
class EmbeddingService:
def __init__(
self,
openai_api_key: str,
model: str = "text-embedding-3-small",
batch_size: int = 10,
max_concurrent_requests: int = 10, # Limit to 10 concurrent requests
):
self.client = AsyncOpenAI(api_key=openai_api_key)
self.model = model
self.batch_size = batch_size
self.semaphore = asyncio.Semaphore(max_concurrent_requests) # Rate limiter
self.total_requests = 0 # Total number of requests to process
self.completed_requests = 0 # Number of completed requests
async def get_embedding(self, text: str) -> List[float]:
"""Generate embeddings for the given text using OpenAI."""
text = text.replace("\n", " ")
try:
async with self.semaphore: # Acquire a semaphore slot
response = await self.client.embeddings.create(
input=[text], model=self.model
)
self.completed_requests += 1 # Increment completed requests
self._log_progress() # Log progress
return response.data[0].embedding
except Exception as e:
logger.error(f"Failed to generate embedding: {e}")
raise OpenAIError(f"OpenAI API error: {e}")
async def create_embeddings(
self,
data: Union[Dataset, List[str]],
target_column: str = None,
output_column: str = "embeddings",
) -> Union[Dataset, List[List[float]]]:
"""
Create embeddings for either a Dataset or a list of strings.
Args:
data: Either a Dataset or a list of strings.
target_column: The column in the Dataset to generate embeddings for (required if data is a Dataset).
output_column: The column to store embeddings in the Dataset (default: "embeddings").
Returns:
If data is a Dataset, returns the Dataset with the embeddings column.
If data is a list of strings, returns a list of embeddings.
"""
if isinstance(data, Dataset):
if not target_column:
raise ValueError("target_column is required when data is a Dataset.")
return await self._create_embeddings_for_dataset(
data, target_column, output_column
)
elif isinstance(data, list):
return await self._create_embeddings_for_texts(data)
else:
raise TypeError(
"data must be either a Hugging Face Dataset or a list of strings."
)
async def _create_embeddings_for_dataset(
self, dataset: Dataset, target_column: str, output_column: str
) -> Dataset:
"""Create embeddings for the target column in the Dataset."""
logger.info("Generating embeddings for Dataset...")
self.total_requests = len(dataset) # Set total number of requests
self.completed_requests = 0 # Reset completed requests counter
embeddings = []
for i in range(0, len(dataset), self.batch_size):
batch = dataset[i : i + self.batch_size]
batch_embeddings = await asyncio.gather(
*[self.get_embedding(text) for text in batch[target_column]]
)
embeddings.extend(batch_embeddings)
dataset = dataset.add_column(output_column, embeddings)
return dataset
async def _create_embeddings_for_texts(self, texts: List[str]) -> List[List[float]]:
"""Create embeddings for a list of strings."""
logger.info("Generating embeddings for list of texts...")
self.total_requests = len(texts) # Set total number of requests
self.completed_requests = 0 # Reset completed requests counter
batches = [
texts[i : i + self.batch_size]
for i in range(0, len(texts), self.batch_size)
]
embeddings = []
for batch in batches:
batch_embeddings = await asyncio.gather(
*[self.get_embedding(text) for text in batch]
)
embeddings.extend(batch_embeddings)
return embeddings
def _log_progress(self):
"""Log the progress of embedding generation."""
progress = (self.completed_requests / self.total_requests) * 100
logger.info(
f"Progress: {self.completed_requests}/{self.total_requests} ({progress:.2f}%)"
)
|