Spaces:
Runtime error
Runtime error
File size: 22,408 Bytes
de38977 82b35ca 30b40e4 82b35ca de38977 82b35ca de38977 30b40e4 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca 30b40e4 82b35ca 30b40e4 82b35ca 30b40e4 82b35ca de38977 82b35ca de38977 30b40e4 82b35ca de38977 82b35ca 30b40e4 82b35ca 30b40e4 82b35ca 30b40e4 82b35ca 30b40e4 82b35ca 30b40e4 82b35ca 30b40e4 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 82b35ca de38977 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 | {
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"id": "cb2ff14b",
"metadata": {},
"outputs": [],
"source": [
"import sys\n",
"sys.path.append('..')"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "6bb3bb7d",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"f:\\Dissertation\\prod-rag-chat\\.venv\\Lib\\site-packages\\tqdm\\auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
" from .autonotebook import tqdm as notebook_tqdm\n"
]
}
],
"source": [
"from datasets import Dataset\n",
"from app import retrieve_relevant_documents, emb_text, model, embedding_model\n",
"from langchain_classic.chains.combine_documents import create_stuff_documents_chain\n",
"from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder\n",
"from langchain_core.runnables import RunnableLambda\n",
"from langchain_core.documents import Document"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "e572fb31",
"metadata": {},
"outputs": [],
"source": [
"def setup_standalone_rag_chain():\n",
" \"\"\"Setup a standalone RAG chain for testing without Chainlit session.\"\"\"\n",
" \n",
" def get_context_and_history(inputs):\n",
" \"\"\"Retrieve context without session history.\"\"\"\n",
" query = inputs[\"question\"]\n",
" relevant_docs = retrieve_relevant_documents(query, limit=5)\n",
" print(\"Relevant documents:\", relevant_docs[0] if relevant_docs else \"No documents found\")\n",
" \n",
" # Convert dictionaries to Document objects for LangChain\n",
" doc_objects = []\n",
" for doc in relevant_docs:\n",
" doc_obj = Document(\n",
" page_content=doc.get('text', ''),\n",
" metadata=doc.get('metadata', {})\n",
" )\n",
" doc_objects.append(doc_obj)\n",
"\n",
" return {\n",
" \"question\": query,\n",
" \"context\": doc_objects,\n",
" \"history\": [] # Empty history for testing\n",
" }\n",
" \n",
" system_prompt = \"\"\"You are a helpful assistant specialising in developing non-discriminatory competence standards and disability support, reasonable adjustments, and equality legislation.\n",
"\n",
"When answering questions, you should:\n",
"1. Use the provided context documents to inform your response\n",
"2. Be accurate and helpful\n",
"3. If the context doesn't contain relevant information, say so clearly\n",
"4. Always reply in English\n",
"5. Provide clear recommendations and examples wherever applicable\n",
"6. Do not make assumptions about the user's knowledge or background\n",
"7. If the user asks for a specific law or regulation, provide a brief explanation and cite relevant documents if available.\n",
"8. Do not overemphasize disability in your responses, but rather focus on the support and adjustments that can be made to ensure equality and inclusivity.\n",
"9. If the user query explicitly asks for a disability-related topic, provide a well-informed response based on the context documents.\n",
"\n",
"Context documents:\n",
"{context} \n",
"\n",
"Please provide a clear response using the above context\n",
"\"\"\"\n",
"\n",
" prompt = ChatPromptTemplate.from_messages([\n",
" (\"system\", system_prompt),\n",
" MessagesPlaceholder(variable_name=\"history\"),\n",
" (\"human\", \"{question}\"),\n",
" ])\n",
"\n",
" question_answer_chain = create_stuff_documents_chain(model, prompt)\n",
" \n",
" # Use a custom chain that properly handles our context and history\n",
" def process_input_and_format(inputs):\n",
" context_data = get_context_and_history(inputs)\n",
" return {\n",
" \"context\": context_data[\"context\"],\n",
" \"question\": context_data[\"question\"],\n",
" \"history\": context_data[\"history\"]\n",
" }\n",
" \n",
" chain = RunnableLambda(process_input_and_format) | question_answer_chain\n",
" \n",
" return chain"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "330ee35d",
"metadata": {},
"outputs": [],
"source": [
"\n",
"# Setup the RAG chain\n",
"rag_chain = setup_standalone_rag_chain()\n",
"\n",
"questions = [\"What are Provisions, Criteria and Practices?\", \n",
" \"What is 'reasonable'?\",\n",
" \"What is 'substantial disadvantage'?\",\n",
" ]\n",
"ground_truths = [\n",
" \"\"\"The Equality and Human Rights Commission (EHRC) interprets PCPs as including:3 \n",
"+ arrangements (for example, for deciding who to admit) \n",
"+ the way that education, or access to any benefit, service or facility is offered or provided \n",
"+ one-off or discretionary decisions \n",
"+ proposals or directions to do something in a particular way \n",
"+ formal and informal policies \n",
"+ rules\"\"\",\n",
"\n",
" \"\"\"There are two key considerations of 'reasonableness' which can help when thinking through \n",
"when an adjustment may be reasonable:4 \n",
"+ Could the adjustment be practicable in its application (is it possible)? \n",
"+ Could the adjustment be effective in achieving its aim (will it work)? \n",
"There is no need to prove that the adjustment is practicable and effective in advance, just \n",
"that it might be. An adjustment should not be considered unreasonable if it does not remove \n",
"the disadvantage fully; an adjustment which partially removes or reduces substantial \n",
"disadvantage is also likely to be reasonable.\"\"\",\n",
"\n",
" \"\"\"'Substantial' is defined in the Act as 'more than minor or trivial'. \n",
"Examples of disadvantage recognised by the EHRC include: \n",
"+ The additional time and effort expended by a disabled student \n",
"+ The inconvenience, indignity, discomfort, or perceived disadvantage suffered by a \n",
"disabled student \n",
"+ The loss of opportunity or diminished progress experienced by a disabled student. \"\"\"]\n",
"\n",
"\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "ba3810dd",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"2025-12-20 19:42:27 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"Relevant documents: {'text': 'What is a provision, criterion or practice? The phrase ‘provision, criterion or practice’ is not defined by the Act. These concepts should be construed widely so as to include, for example, any formal or informal policies, rules, practices, arrangements, criteria, procedures, activities or provisions. They can cover one-off decisions and actions. In simple terms, they are about the way an education provider does things. Example:', 'metadata': {'source': 'data\\\\technical-guidance-further-higher-education.docx', 'file_directory': 'data', 'filename': 'technical-guidance-further-higher-education.docx', 'last_modified': '2025-07-02T21:00:50', 'page_number': 95, 'languages': ['eng'], 'filetype': 'application/vnd.openxmlformats-officedocument.wordprocessingml.document', 'category': 'CompositeElement', 'element_id': '3ae881ad6f81487213a9e234debf0921'}, 'score': 0.7807720899581909}\n",
"2025-12-20 19:42:29 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:32 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:32 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"Relevant documents: {'text': '‘Reasonable’ means having regard to all of the circumstances including the nature of the act and how obviously discriminatory it is, the authority of the person making the statement and the knowledge that the helper has or ought to have.', 'metadata': {'source': 'data\\\\technical-guidance-further-higher-education.docx', 'file_directory': 'data', 'filename': 'technical-guidance-further-higher-education.docx', 'last_modified': '2025-07-02T21:00:50', 'page_number': 36, 'languages': ['eng'], 'filetype': 'application/vnd.openxmlformats-officedocument.wordprocessingml.document', 'category': 'CompositeElement', 'element_id': 'c5e3a60e2a6ccc88e0eff961f645a962'}, 'score': 0.7367081046104431}\n",
"2025-12-20 19:42:33 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:36 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:37 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"Relevant documents: {'text': 'The Act states that disadvantage must be substantial, which is defined as more than minor or trivial. Whether such a disadvantage exists in a particular case is a question of fact, and is assessed on an objective basis. s212(1)', 'metadata': {'source': 'data\\\\technical-guidance-further-higher-education.docx', 'file_directory': 'data', 'filename': 'technical-guidance-further-higher-education.docx', 'last_modified': '2025-07-02T21:00:50', 'page_number': 89, 'languages': ['eng'], 'filetype': 'application/vnd.openxmlformats-officedocument.wordprocessingml.document', 'category': 'CompositeElement', 'element_id': 'b9e8ef04daf9150c9f7e32736b53df5b'}, 'score': 0.8376985788345337}\n",
"2025-12-20 19:42:38 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:40 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n"
]
}
],
"source": [
"answers = []\n",
"contexts = []\n",
"\n",
"def clean_answer(answer):\n",
" \"\"\"Remove <think></think> tags and content from the answer.\"\"\"\n",
" import re\n",
" # Remove everything between <think> and </think> tags, including the tags themselves\n",
" cleaned = re.sub(r'<think>.*?</think>\\s*', '', answer, flags=re.DOTALL)\n",
" return cleaned.strip()\n",
"\n",
"# Inference\n",
"for query in questions:\n",
" # Get answer from the RAG chain\n",
" answer = rag_chain.invoke({\"question\": query})\n",
" # Clean the answer to remove thinking content\n",
" cleaned_answer = clean_answer(answer)\n",
" answers.append(cleaned_answer)\n",
" \n",
" # Get relevant documents for context\n",
" relevant_docs = retrieve_relevant_documents(query, limit=5)\n",
" context_texts = [doc['text'] for doc in relevant_docs]\n",
" contexts.append(context_texts)\n",
"\n",
"# To dict\n",
"data = {\n",
" \"question\": questions,\n",
" \"answer\": answers,\n",
" \"contexts\": contexts,\n",
" \"reference\": ground_truths\n",
"}\n",
"\n",
"# Convert dict to dataset\n",
"dataset = Dataset.from_dict(data)\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "25c00c40",
"metadata": {},
"outputs": [],
"source": [
"from langchain_nebius import ChatNebius\n",
"from pydantic import SecretStr\n",
"import os"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "2994b26e",
"metadata": {},
"outputs": [],
"source": [
"eval_model = ChatNebius(\n",
" model=\"Qwen/Qwen3-235B-A22B-Instruct-2507\",\n",
" streaming=False, # <--- CRITICAL: Must be False for Ragas\n",
" temperature=0, # Lower temperature for the judge is better\n",
" max_tokens=8192,\n",
" top_p=0.95,\n",
" api_key=SecretStr(os.getenv(\"OPENAI_API_KEY\")),\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "3e016be2",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Evaluating: 0%| | 0/12 [00:00<?, ?it/s]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"2025-12-20 19:42:46 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:48 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:49 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:49 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:49 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Evaluating: 8%|▊ | 1/12 [00:06<01:08, 6.19s/it]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"2025-12-20 19:42:50 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:50 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:50 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:50 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:50 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:50 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:50 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:55 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:56 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:57 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:57 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:42:59 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:00 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:00 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:01 - HTTP Request: POST https://api.studio.nebius.ai/v1/embeddings \"HTTP/1.1 200 OK\"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Evaluating: 17%|█▋ | 2/12 [00:20<01:49, 10.98s/it]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"2025-12-20 19:43:05 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:05 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:05 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:05 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Evaluating: 42%|████▏ | 5/12 [00:26<00:32, 4.60s/it]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"2025-12-20 19:43:11 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:11 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Evaluating: 50%|█████ | 6/12 [00:28<00:22, 3.83s/it]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"2025-12-20 19:43:17 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:18 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:18 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:22 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:22 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Evaluating: 58%|█████▊ | 7/12 [00:41<00:31, 6.21s/it]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"2025-12-20 19:43:25 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:25 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:25 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:28 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Evaluating: 67%|██████▋ | 8/12 [00:45<00:23, 5.78s/it]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"2025-12-20 19:43:31 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:31 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n",
"2025-12-20 19:43:31 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Evaluating: 75%|███████▌ | 9/12 [00:52<00:17, 5.95s/it]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"2025-12-20 19:43:36 - HTTP Request: POST https://api.studio.nebius.ai/v1/chat/completions \"HTTP/1.1 200 OK\"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Evaluating: 100%|██████████| 12/12 [00:53<00:00, 4.47s/it]\n"
]
}
],
"source": [
"from ragas import evaluate\n",
"from ragas.metrics import (\n",
" faithfulness,\n",
" answer_relevancy,\n",
" context_recall,\n",
" context_precision,\n",
")\n",
"\n",
"result = evaluate(\n",
" llm=eval_model,\n",
" embeddings=embedding_model,\n",
" dataset = dataset, \n",
" metrics=[\n",
" context_precision,\n",
" context_recall,\n",
" faithfulness,\n",
" answer_relevancy,\n",
" ],\n",
")\n",
"\n",
"df = result.to_pandas()\n",
"\n",
"# evaluation_results = result.to_pandas()\n",
"\n",
"# display_columns = ['user_input', 'answer_relevancy', 'faithfulness', 'context_precision', 'context_recall']\n",
"# formatted_results = evaluation_results[display_columns].to_markdown(index=False, numalign=\"left\", stralign=\"left\")\n",
"\n",
"# print(formatted_results)\n"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "d8514ff3",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" user_input \\\n",
"0 What are Provisions, Criteria and Practices? \n",
"1 What is 'reasonable'? \n",
"2 What is 'substantial disadvantage'? \n",
"\n",
" retrieved_contexts \\\n",
"0 [What is a provision, criterion or practice? T... \n",
"1 [‘Reasonable’ means having regard to all of th... \n",
"2 [The Act states that disadvantage must be subs... \n",
"\n",
" response \\\n",
"0 **Provisions, Criteria and Practices (PCPs)** ... \n",
"1 The term **'reasonable'** in the context of di... \n",
"2 'Substantial disadvantage' refers to a disadva... \n",
"\n",
" reference context_precision \\\n",
"0 The Equality and Human Rights Commission (EHRC... 1.000000 \n",
"1 There are two key considerations of 'reasonabl... 0.533333 \n",
"2 'Substantial' is defined in the Act as 'more t... 0.866667 \n",
"\n",
" context_recall faithfulness answer_relevancy \n",
"0 0.857143 0.714286 0.704690 \n",
"1 0.333333 0.774194 0.716996 \n",
"2 0.400000 0.875000 0.834463 \n"
]
}
],
"source": [
"print(df)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": ".venv",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.12.5"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
|