File size: 10,916 Bytes
bc61879 da1d68a 7a54e0d 3fbe6fe da1d68a 3fbe6fe da1d68a bc61879 7a54e0d 3fbe6fe 7a54e0d da1d68a 7a54e0d da1d68a 3fbe6fe 7a54e0d bc61879 7a54e0d bc61879 1351a87 bc61879 1351a87 bc61879 1351a87 bc61879 1351a87 bc61879 1351a87 bc61879 1351a87 bc61879 da3f47f bc61879 da3f47f bc61879 da3f47f bc61879 da3f47f bc61879 da3f47f da1d68a bc61879 da1d68a bc61879 da3f47f bc61879 1351a87 bc61879 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 |
import os
from datetime import datetime
import json
from huggingface_hub import HfApi
import gradio as gr
import csv
import pandas as pd
import io
from typing import TypedDict, List
from climateqa.constants import DOCUMENT_METADATA_DEFAULT_VALUES
from langchain_core.documents import Document
def serialize_docs(docs:list[Document])->list:
"""Convert document objects to a simplified format compatible with Hugging Face datasets.
This function processes document objects by extracting their page content and metadata,
normalizing the metadata structure to ensure consistency. It applies default values
from DOCUMENT_METADATA_DEFAULT_VALUES for any missing metadata fields.
Args:
docs (list): List of document objects, each with page_content and metadata attributes
Returns:
list: List of dictionaries with standardized "page_content" and "metadata" fields
"""
new_docs = []
for doc in docs:
# Make sure we have a clean doc format
new_doc = {
"page_content": doc.page_content,
"metadata": {}
}
# Ensure all metadata fields exist with defaults if missing
for field, default_value in DOCUMENT_METADATA_DEFAULT_VALUES.items():
new_value = doc.metadata.get(field, default_value)
try:
new_doc["metadata"][field] = type(default_value)(new_value)
except:
new_doc["metadata"][field] = default_value
new_docs.append(new_doc)
if new_docs == []:
new_docs = [{"page_content": "No documents found", "metadata": DOCUMENT_METADATA_DEFAULT_VALUES}]
return new_docs
## AZURE LOGGING - DEPRECATED
def log_on_azure(file, logs, share_client):
"""Log data to Azure Blob Storage.
Args:
file (str): Name of the file to store logs
logs (dict): Log data to store
share_client: Azure share client instance
"""
logs = json.dumps(logs)
file_client = share_client.get_file_client(file)
file_client.upload_file(logs)
def log_interaction_to_azure(history, output_query, sources, docs, share_client, user_id):
"""Log chat interaction to Azure and Hugging Face.
Args:
history (list): Chat message history
output_query (str): Processed query
sources (list): Knowledge base sources used
docs (list): Retrieved documents
share_client: Azure share client instance
user_id (str): User identifier
"""
try:
# Log interaction to Azure if not in local environment
if os.getenv("GRADIO_ENV") != "local":
timestamp = str(datetime.now().timestamp())
prompt = history[1]["content"]
logs = {
"user_id": str(user_id),
"prompt": prompt,
"query": prompt,
"question": output_query,
"sources": sources,
"docs": serialize_docs(docs),
"answer": history[-1].content,
"time": timestamp,
}
# Log to Azure
log_on_azure(f"{timestamp}.json", logs, share_client)
except Exception as e:
print(f"Error logging on Azure Blob Storage: {e}")
error_msg = f"ClimateQ&A Error: {str(e)[:100]} - The error has been noted, try another question and if the error remains, you can contact us :)"
raise gr.Error(error_msg)
def log_drias_interaction_to_azure(query, sql_query, data, share_client, user_id):
"""Log Drias data interaction to Azure and Hugging Face.
Args:
query (str): User query
sql_query (str): SQL query used
data: Retrieved data
share_client: Azure share client instance
user_id (str): User identifier
"""
try:
# Log interaction to Azure if not in local environment
if os.getenv("GRADIO_ENV") != "local":
timestamp = str(datetime.now().timestamp())
logs = {
"user_id": str(user_id),
"query": query,
"sql_query": sql_query,
"time": timestamp,
}
log_on_azure(f"drias_{timestamp}.json", logs, share_client)
print(f"Logged Drias interaction to Azure Blob Storage: {logs}")
else:
print("share_client or user_id is None, or GRADIO_ENV is local")
except Exception as e:
print(f"Error logging Drias interaction on Azure Blob Storage: {e}")
error_msg = f"Drias Error: {str(e)[:100]} - The error has been noted, try another question and if the error remains, you can contact us :)"
raise gr.Error(error_msg)
## HUGGING FACE LOGGING
def log_on_huggingface(log_filename, logs, log_type="chat"):
"""Log data to Hugging Face dataset repository.
Args:
log_filename (str): Name of the file to store logs
logs (dict): Log data to store
log_type (str): Type of log to store
"""
try:
if log_type =="chat":
# Get Hugging Face token from environment
hf_token = os.getenv("HF_LOGS_TOKEN")
if not hf_token:
print("HF_LOGS_TOKEN not found in environment variables")
return
# Get repository name from environment or use default
repo_id = os.getenv("HF_DATASET_REPO", "Ekimetrics/climateqa_logs")
elif log_type =="drias":
# Get Hugging Face token from environment
hf_token = os.getenv("HF_LOGS_DRIAS_TOKEN")
if not hf_token:
print("HF_LOGS_DRIAS_TOKEN not found in environment variables")
return
# Get repository name from environment or use default
repo_id = os.getenv("HF_DATASET_REPO_DRIAS", "Ekimetrics/climateqa_logs_talk_to_data")
else:
raise ValueError(f"Invalid log type: {log_type}")
# Initialize HfApi
api = HfApi(token=hf_token)
# Add timestamp to the log data
logs["timestamp"] = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
# Convert logs to JSON string
logs_json = json.dumps(logs)
# Upload directly from memory
api.upload_file(
path_or_fileobj=logs_json.encode('utf-8'),
path_in_repo=log_filename,
repo_id=repo_id,
repo_type="dataset"
)
except Exception as e:
print(f"Error logging to Hugging Face: {e}")
def log_interaction_to_huggingface(history, output_query, sources, docs, share_client, user_id):
"""Log chat interaction to Hugging Face.
Args:
history (list): Chat message history
output_query (str): Processed query
sources (list): Knowledge base sources used
docs (list): Retrieved documents
share_client: Azure share client instance (unused in this function)
user_id (str): User identifier
"""
try:
# Log interaction if not in local environment
if os.getenv("GRADIO_ENV") != "local":
timestamp = str(datetime.now().timestamp())
prompt = history[1]["content"]
logs = {
"user_id": str(user_id),
"prompt": prompt,
"query": prompt,
"question": output_query,
"sources": sources,
"docs": serialize_docs(docs),
"answer": history[-1].content,
"time": timestamp,
}
# Log to Hugging Face
log_on_huggingface(f"chat/{timestamp}.json", logs, log_type="chat")
print(f"Logged interaction to Hugging Face")
else:
print("Did not log to Hugging Face because GRADIO_ENV is local")
except Exception as e:
print(f"Error logging to Hugging Face: {e}")
error_msg = f"ClimateQ&A Error: {str(e)[:100]})"
raise gr.Error(error_msg)
def log_drias_interaction_to_huggingface(query, sql_query, user_id):
"""Log Drias data interaction to Hugging Face.
Args:
query (str): User query
sql_query (str): SQL query used
data: Retrieved data
user_id (str): User identifier
"""
try:
if os.getenv("GRADIO_ENV") != "local":
timestamp = str(datetime.now().timestamp())
logs = {
"user_id": str(user_id),
"query": query,
"sql_query": sql_query,
"time": timestamp,
}
log_on_huggingface(f"drias/drias_{timestamp}.json", logs, log_type="drias")
print(f"Logged Drias interaction to Hugging Face: {logs}")
else:
print("share_client or user_id is None, or GRADIO_ENV is local")
except Exception as e:
print(f"Error logging Drias interaction to Hugging Face: {e}")
error_msg = f"Drias Error: {str(e)[:100]} - The error has been noted, try another question and if the error remains, you can contact us :)"
raise gr.Error(error_msg)
def log_interaction(history, output_query, sources, docs, share_client, user_id):
"""Log chat interaction to Hugging Face, and fall back to Azure if that fails.
Args:
history (list): Chat message history
output_query (str): Processed query
sources (list): Knowledge base sources used
docs (list): Retrieved documents
share_client: Azure share client instance
user_id (str): User identifier
"""
try:
# First try to log to Hugging Face
log_interaction_to_huggingface(history, output_query, sources, docs, share_client, user_id)
except Exception as e:
print(f"Failed to log to Hugging Face, falling back to Azure: {e}")
try:
# Fall back to Azure logging
if os.getenv("GRADIO_ENV") != "local":
timestamp = str(datetime.now().timestamp())
prompt = history[1]["content"]
logs = {
"user_id": str(user_id),
"prompt": prompt,
"query": prompt,
"question": output_query,
"sources": sources,
"docs": serialize_docs(docs),
"answer": history[-1].content,
"time": timestamp,
}
# Log to Azure
log_on_azure(f"{timestamp}.json", logs, share_client)
print("Successfully logged to Azure as fallback")
except Exception as azure_error:
print(f"Error in Azure fallback logging: {azure_error}")
error_msg = f"ClimateQ&A Logging Error: {str(azure_error)[:100]})"
# Don't raise error to avoid disrupting user experience
print(error_msg)
|