File size: 21,672 Bytes
488dc3e
1bbca12
4701375
 
 
488dc3e
4701375
488dc3e
4701375
 
 
 
 
 
1bbca12
4701375
 
 
 
 
 
 
488dc3e
4701375
 
488dc3e
 
 
 
1bbca12
 
 
4701375
 
 
 
1bbca12
488dc3e
4701375
 
 
 
 
1bbca12
4701375
488dc3e
4701375
 
 
 
1bbca12
4701375
488dc3e
4701375
488dc3e
 
 
4701375
 
488dc3e
4701375
 
488dc3e
 
4701375
 
488dc3e
4701375
 
 
 
 
 
 
 
 
 
 
488dc3e
4701375
 
488dc3e
 
4701375
 
488dc3e
 
4701375
 
 
488dc3e
4701375
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
488dc3e
4701375
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
488dc3e
4701375
 
 
 
488dc3e
 
4701375
 
488dc3e
 
 
4701375
488dc3e
 
4701375
 
 
 
 
 
 
488dc3e
4701375
488dc3e
4701375
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
488dc3e
4701375
 
 
 
488dc3e
 
4701375
488dc3e
 
4701375
 
488dc3e
4701375
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
488dc3e
4701375
 
488dc3e
 
4701375
488dc3e
 
 
 
 
 
 
 
4701375
488dc3e
 
 
 
 
 
 
 
 
 
 
4701375
488dc3e
4701375
488dc3e
 
 
 
 
4701375
488dc3e
4701375
488dc3e
 
4701375
488dc3e
 
4701375
 
488dc3e
 
4701375
488dc3e
 
 
 
4701375
488dc3e
4701375
488dc3e
 
 
 
4701375
488dc3e
 
 
 
 
4701375
488dc3e
 
 
4701375
 
 
 
488dc3e
 
 
 
4701375
 
 
 
 
 
 
488dc3e
 
4701375
488dc3e
 
4701375
488dc3e
4701375
488dc3e
4701375
 
 
 
 
488dc3e
4701375
488dc3e
4701375
488dc3e
 
4701375
488dc3e
 
 
4701375
 
 
488dc3e
 
 
 
4701375
488dc3e
 
 
 
 
 
 
 
4701375
 
488dc3e
 
 
 
 
 
 
4701375
488dc3e
 
 
 
4701375
488dc3e
 
 
 
 
 
4701375
488dc3e
 
 
4701375
 
488dc3e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4701375
488dc3e
4701375
488dc3e
 
 
 
4701375
 
488dc3e
 
 
4701375
488dc3e
 
 
 
 
 
 
 
4701375
488dc3e
 
 
 
1bbca12
 
4701375
1bbca12
488dc3e
4701375
488dc3e
 
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
import os
import json
import logging
import asyncio
import aiohttp
import nest_asyncio
import requests
import pandas as pd
from typing import Dict, Any, List
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.messages import SystemMessage, HumanMessage
from langgraph.graph import StateGraph, END
from sentence_transformers import SentenceTransformer
import gradio as gr
from dotenv import load_dotenv
from huggingface_hub import InferenceClient
from state import JARVISState
from tools import (
    search_tool, multi_hop_search_tool, file_parser_tool, image_parser_tool,
    calculator_tool, document_retriever_tool, duckduckgo_search_tool,
    weather_info_tool, hub_stats_tool, guest_info_retriever_tool
)

# Setup logging
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
logger = logging.getLogger(__name__)

# Apply nest_asyncio
nest_asyncio.apply()

# Load environment variables
load_dotenv()
SPACE_ID = os.getenv("SPACE_ID", "onisj/jarvis_gaia_agent")
GAIA_API_URL = "https://agents-course-unit4-scoring.hf.space"
GAIA_FILE_URL = f"{GAIA_API_URL}/files/"
HF_TOKEN = os.getenv("HUGGINGFACEHUB_API_TOKEN")

# Verify environment variables
if not SPACE_ID:
    raise ValueError("SPACE_ID not set")
if not HF_TOKEN:
    raise ValueError("HUGGINGFACEHUB_API_TOKEN not set")
logger.info(f"SPACE_ID: {SPACE_ID}")

# Initialize models
try:
    llm = InferenceClient(
        model="meta-llama/Meta-Llama-3-8B-Instruct",
        token=HF_TOKEN,
        timeout=30
    )
    logger.info("Hugging Face Inference LLM initialized")
except Exception as e:
    logger.error(f"Failed to initialize LLM: {e}")
    llm = None

try:
    embedder = SentenceTransformer("all-MiniLM-L6-v2")
    logger.info("Sentence transformer initialized")
except Exception as e:
    logger.error(f"Failed to initialize embedder: {e}")
    embedder = None

# --- Helper Functions ---
async def test_gaia_api(task_id: str, file_type: str = "txt") -> tuple[bool, str | None]:
    """Test if a file exists for the task ID."""
    try:
        for ext in [file_type, "txt", "csv", "xlsx", "jpg", "pdf"]:
            async with aiohttp.ClientSession() as session:
                async with session.get(f"{GAIA_FILE_URL}{task_id}.{ext}", timeout=5) as resp:
                    logger.info(f"GAIA API test for task {task_id} with .{ext}: HTTP {resp.status}")
                    if resp.status == 200:
                        file_path = f"temp_{task_id}.{ext}"
                        with open(file_path, "wb") as f:
                            f.write(await resp.read())
                        return True, ext
        logger.info(f"No file found for task {task_id}")
        return False, None
    except Exception as e:
        logger.warning(f"GAIA API test failed: {str(e)}")
        return False, None

# --- Node Functions ---
async def parse_question(state: Dict[str, Any]) -> Dict[str, Any]:
    """Parse the question to select appropriate tools."""
    try:
        question = state["question"]
        task_id = state["task_id"]
        tools_needed = ["search_tool"]

        if llm:
            prompt = ChatPromptTemplate.from_messages([
                SystemMessage(content="""Select tools from: ['search_tool', 'multi_hop_search_tool', 'file_parser_tool', 'image_parser_tool', 'calculator_tool', 'document_retriever_tool', 'duckduckgo_search_tool', 'weather_info_tool', 'hub_stats_tool', 'guest_info_retriever_tool'].
                Return JSON list, e.g., ["search_tool", "file_parser_tool"].
                Rules:
                - Always include "search_tool" unless purely computational.
                - Use "multi_hop_search_tool" for complex queries (over 20 words).
                - Use "file_parser_tool" for data, tables, or Excel.
                - Use "image_parser_tool" for images/videos.
                - Use "calculator_tool" for math calculations.
                - Use "document_retriever_tool" for documents/PDFs.
                - Use "duckduckgo_search_tool" for additional search capability.
                - Use "weather_info_tool" for weather-related queries.
                - Use "hub_stats_tool" for Hugging Face Hub queries.
                - Use "guest_info_retriever_tool" for guest-related queries.
                - Output ONLY valid JSON."""),
                HumanMessage(content=f"Query: {question}")
            ])
            try:
                response = llm.chat_completion(
                    messages=[
                        {"role": "system", "content": prompt[0].content},
                        {"role": "user", "content": prompt[1].content}
                    ],
                    max_tokens=512,
                    temperature=0.7
                )
                tools_needed = json.loads(response["choices"][0]["message"]["content"].strip())
                valid_tools = {
                    "search_tool", "multi_hop_search_tool", "file_parser_tool", "image_parser_tool",
                    "calculator_tool", "document_retriever_tool", "duckduckgo_search_tool",
                    "weather_info_tool", "hub_stats_tool", "guest_info_retriever_tool"
                }
                tools_needed = [tool for tool in tools_needed if tool in valid_tools]
            except Exception as e:
                logger.warning(f"Task {task_id} failed: JSON parse error: {e}")
                tools_needed = ["search_tool"]

        # Keyword-based fallback
        question_lower = question.lower()
        if any(word in question_lower for word in ["image", "video"]):
            tools_needed.append("image_parser_tool")
        if any(word in question_lower for word in ["data", "table", "excel"]):
            tools_needed.append("file_parser_tool")
        if any(word in question_lower for word in ["calculate", "math"]):
            tools_needed.append("calculator_tool")
        if any(word in question_lower for word in ["document", "pdf"]):
            tools_needed.append("document_retriever_tool")
        if any(word in question_lower for word in ["weather"]):
            tools_needed.append("weather_info_tool")
        if any(word in question_lower for word in ["model", "huggingface"]):
            tools_needed.append("hub_stats_tool")
        if any(word in question_lower for word in ["guest", "name", "relation"]):
            tools_needed.append("guest_info_retriever_tool")
        if len(question.split()) > 20:
            tools_needed.append("multi_hop_search_tool")

        file_available, file_ext = await test_gaia_api(task_id)
        if file_available:
            if "file_parser_tool" not in tools_needed and any(word in question_lower for word in ["data", "table", "excel"]):
                tools_needed.append("file_parser_tool")
            if "image_parser_tool" not in tools_needed and "image" in question_lower:
                tools_needed.append("image_parser_tool")
            if "document_retriever_tool" not in tools_needed and file_ext == "pdf":
                tools_needed.append("document_retriever_tool")
        else:
            tools_needed = [tool for tool in tools_needed if tool not in ["file_parser_tool", "image_parser_tool", "document_retriever_tool"]]

        state["tools_needed"] = list(set(tools_needed))  # Remove duplicates
        logger.info(f"Task {task_id}: Selected tools: {tools_needed}")
        return state
    except Exception as e:
        logger.error(f"Error parsing task {task_id}: {e}")
        state["tools_needed"] = ["search_tool"]
        return state

async def tool_dispatcher(state: JARVISState) -> JARVISState:
    """Dispatch selected tools to process the state."""
    try:
        updated_state = state.copy()
        file_type = "jpg" if "image" in state["question"].lower() else "txt"
        if "menu" in state["question"].lower() or "report" in state["question"].lower():
            file_type = "pdf"
        elif "data" in state["question"].lower():
            file_type = "xlsx"

        can_download, file_ext = await test_gaia_api(updated_state["task_id"], file_type)

        for tool in updated_state["tools_needed"]:
            try:
                if tool == "search_tool":
                    result = await search_tool.ainvoke({"query": updated_state["question"]})
                    updated_state["web_results"].extend([r["content"] for r in result])
                elif tool == "multi_hop_search_tool":
                    result = await multi_hop_search_tool.ainvoke({"query": updated_state["question"], "steps": 3})
                    updated_state["web_results"].extend([r["content"] for r in result])
                    await asyncio.sleep(2)  # Rate limit
                elif tool == "file_parser_tool" and can_download:
                    result = await file_parser_tool.ainvoke({"task_id": updated_state["task_id"], "file_type": file_ext})
                    updated_state["file_results"] = str(result)
                elif tool == "image_parser_tool" and can_download:
                    result = await image_parser_tool.ainvoke({
                        "file_path": f"temp_{updated_state['task_id']}.{file_ext}",
                        "task": "describe"
                    })
                    updated_state["image_results"] = str(result)
                elif tool == "calculator_tool":
                    result = await calculator_tool.ainvoke({"expression": updated_state.get("question", "")})
                    updated_state["calculation_results"] = str(result)
                elif tool == "document_retriever_tool" and can_download:
                    result = await document_retriever_tool.ainvoke({
                        "task_id": updated_state["task_id"],
                        "query": updated_state["question"],
                        "file_type": file_ext
                    })
                    updated_state["document_results"] = str(result)
                elif tool == "duckduckgo_search_tool":
                    result = await duckduckgo_search_tool.run(updated_state["question"])
                    updated_state["web_results"].append(str(result))
                elif tool == "weather_info_tool":
                    location = updated_state["question"].split("weather in ")[1].split()[0] if "weather in" in updated_state["question"].lower() else "Unknown"
                    result = await weather_info_tool.ainvoke({"location": location})
                    updated_state["web_results"].append(str(result))
                elif tool == "hub_stats_tool":
                    author = updated_state["question"].split("by ")[1].split()[0] if "by" in updated_state["question"].lower() else "Unknown"
                    result = await hub_stats_tool.ainvoke({"author": author})
                    updated_state["web_results"].append(str(result))
                elif tool == "guest_info_retriever_tool":
                    query = updated_state["question"].split("about ")[1] if "about" in updated_state["question"].lower() else updated_state["question"]
                    result = await guest_info_retriever_tool.ainvoke({"query": query})
                    updated_state["web_results"].append(str(result))
            except Exception as e:
                logger.warning(f"Error in tool {tool} for task {updated_state['task_id']}: {str(e)}")
                updated_state[f"{tool}_results"] = f"Error: {str(e)}"

        logger.info(f"Task {updated_state['task_id']}: Tool results: {updated_state}")
        return updated_state
    except Exception as e:
        logger.error(f"Tool dispatch failed for task {state['task_id']}: {e}")
        return state

async def reasoning(state: JARVISState) -> Dict[str, Any]:
    """Generate exact-match answer with specific formatting."""
    try:
        if not llm:
            return {"answer": "LLM unavailable"}
        prompt = ChatPromptTemplate.from_messages([
            SystemMessage(content="""Provide ONLY the exact answer (e.g., '90', 'HUE'). For USD, use two decimal places (e.g., '1234.00'). For lists, use comma-separated values (e.g., 'Smith, Lee'). For IOC codes, use three-letter codes (e.g., 'ARG'). No explanations or conversational text."""),
            HumanMessage(content="""Question: {question}
Web results: {web_results}
File results: {file_results}
Image results: {image_results}
Calculation results: {calculation_results}
Document results: {document_results}""")
        ])
        response = llm.chat_completion(
            messages=[
                {"role": "system", "content": prompt[0].content},
                {"role": "user", "content": prompt[1].content.format(
                    question=state["question"],
                    web_results="\n".join(state["web_results"]),
                    file_results=state["file_results"],
                    image_results=state["image_results"],
                    calculation_results=state["calculation_results"],
                    document_results=state["document_results"]
                )}
            ],
            max_tokens=512,
            temperature=0.7
        )
        answer = response["choices"][0]["message"]["content"].strip()
        # Clean answer for specific formats
        if "USD" in state["question"].lower():
            try:
                answer = f"{float(answer):.2f}"
            except ValueError:
                pass
        if "before and after" in state["question"].lower():
            answer = answer.replace(" and ", ", ")
        elif "IOC code" in state["question"].lower():
            answer = answer.upper()[:3]
        logger.info(f"Task {state['task_id']}: Answer: {answer}")
        return {"answer": answer}
    except Exception as e:
        logger.error(f"Reasoning failed for task {state['task_id']}: {e}")
        return {"answer": f"Error: {str(e)}"}

def router(state: JARVISState) -> str:
    """Route based on tools needed."""
    if state["tools_needed"]:
        return "tool_dispatcher"
    return "reasoning"

# --- Define StateGraph ---
workflow = StateGraph(JARVISState)
workflow.add_node("parse", parse_question)
workflow.add_node("tool_dispatcher", tool_dispatcher)
workflow.add_node("reasoning", reasoning)
workflow.set_entry_point("parse")
workflow.add_conditional_edges(
    "parse",
    router,
    {
        "tool_dispatcher": "tool_dispatcher",
        "reasoning": "reasoning"
    }
)
workflow.add_edge("tool_dispatcher", "reasoning")
workflow.add_edge("reasoning", END)
graph = workflow.compile()

# --- Basic Agent ---
class BasicAgent:
    def __init__(self):
        logger.info("BasicAgent initialized.")

    async def process_question(self, task_id: str, question: str) -> str:
        """Process a single question with file handling."""
        file_type = "jpg" if "image" in question.lower() else "txt"
        if "menu" in question.lower() or "report" in question.lower():
            file_type = "pdf"
        elif "data" in question.lower():
            file_type = "xlsx"

        file_path = f"temp_{task_id}.{file_type}"
        file_available, file_ext = await test_gaia_api(task_id, file_type)
        if file_available:
            try:
                async with aiohttp.ClientSession() as session:
                    async with session.get(f"{GAIA_FILE_URL}{task_id}.{file_ext}") as resp:
                        if resp.status == 200:
                            with open(file_path, "wb") as f:
                                f.write(await resp.read())
                        else:
                            logger.warning(f"Failed to fetch file for {task_id}: HTTP {resp.status}")
            except Exception as e:
                logger.error(f"Error downloading file for task {task_id}: {str(e)}")

        state = JARVISState(
            task_id=task_id,
            question=question,
            tools_needed=["search_tool"],
            web_results=[],
            file_results="",
            image_results="",
            calculation_results="",
            document_results="",
            messages=[HumanMessage(content=question)],
            answer=""
        )
        try:
            result = await graph.ainvoke(state)
            answer = result["answer"] or "Unknown"
            logger.info(f"Task {task_id}: Final answer generated: {answer}")
            return answer
        except Exception as e:
            logger.error(f"Error processing task {task_id}: {e}")
            return f"Error: {str(e)}"
        finally:
            for ext in ["txt", "csv", "xlsx", "jpg", "pdf"]:
                file_path = f"temp_{task_id}.{ext}"
                if os.path.exists(file_path):
                    try:
                        os.remove(file_path)
                    except Exception as e:
                        logger.error(f"Error removing file {file_path}: {e}")

    async def async_call(self, question: str, task_id: str) -> str:
        return await self.process_question(question, task_id)

    def __call__(self, question: str, task_id: str = None) -> str:
        logger.info(f"Processing question: {question[:50]}...")
        if task_id is None:
            task_id = "unknown_task_id"
        try:
            loop = asyncio.get_event_loop()
        except RuntimeError:
            loop = asyncio.new_event_loop()
            asyncio.set_event_loop(loop)
        return loop.run_until_complete(self.async_call(question, task_id))

# --- Evaluation and Submission ---
def run_and_submit_all(profile: gr.OAuthProfile | None):
    """Run evaluation and submit answers to GAIA API."""
    if not profile:
        logger.error("User not logged in.")
        return "Please Login to Hugging Face.", None
    username = f"{profile.username}"
    logger.info(f"User logged in: {username}")

    questions_url = f"{GAIA_API_URL}/questions"
    submit_url = f"{GAIA_API_URL}/submit"
    agent_code = f"https://huggingface.co/spaces/{SPACE_ID}/tree/main"

    try:
        agent = BasicAgent()
    except Exception as e:
        logger.error(f"Agent initialization failed: {e}")
        return f"Error initializing agent: {e}", None

    logger.info(f"Fetching questions from: {questions_url}")
    try:
        response = requests.get(questions_url, timeout=15)
        response.raise_for_status()
        questions_data = response.json()
        if not questions_data:
            logger.error("Empty questions list.")
            return "No questions fetched.", None
        logger.info(f"Fetched {len(questions_data)} questions.")
    except Exception as e:
        logger.error(f"Error fetching questions: {e}")
        return f"Error fetching questions: {e}", None

    results_log = []
    answers_payload = []
    logger.info(f"Processing {len(questions_data)} questions...")
    for item in questions_data:
        task_id = item.get("task_id")
        question_text = item.get("question")
        if not task_id or question_text is None:
            logger.warning(f"Skipping invalid item: {item}")
            continue
        try:
            submitted_answer = agent(question_text, task_id)
            answers_payload.append({"task_id": task_id, "submitted_answer": submitted_answer})
            results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": submitted_answer})
        except Exception as e:
            logger.error(f"Error for task {task_id}: {e}")
            results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": f"AGENT ERROR: {e}"})

    if not answers_payload:
        logger.error("No answers generated.")
        return "No answers to submit.", pd.DataFrame(results_log)

    submission_data = {"username": username.strip(), "agent_code": agent_code, "answers": answers_payload}
    logger.info(f"Submitting {len(answers_payload)} answers to: {submit_url}")
    try:
        response = requests.post(submit_url, json=submission_data, timeout=120)
        response.raise_for_status()
        result_data = response.json()
        final_status = (
            f"Submission Successful!\n"
            f"User: {result_data.get('username')}\n"
            f"Overall Score: {result_data.get('score', 'N/A')}% "
            f"({result_data.get('correct_count', '?')}/{result_data.get('total_attempted', '?')} correct)\n"
            f"Message: {result_data.get('message', 'No message received.')}"
        )
        results_df = pd.DataFrame(results_log)
        return final_status, results_df
    except Exception as e:
        logger.error(f"Submission failed: {e}")
        results_df = pd.DataFrame(results_log)
        return f"Submission Failed: {e}", results_df

# --- Gradio Interface ---
with gr.Blocks() as demo:
    gr.Markdown("# Evolved JARVIS Agent Evaluation")
    gr.Markdown(
        """
        **Instructions:**

        1. Log in to Hugging Face using the button below.
        2. Click 'Run Evaluation & Submit All Answers' to process GAIA questions and submit.

        ---
        **Disclaimers:**
        Uses Hugging Face Inference, SERPAPI, and OpenWeatherMap for GAIA benchmark.
        """
    )

    gr.LoginButton()

    run_button = gr.Button("Run Evaluation & Submit All Answers")

    status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
    results_table = gr.DataFrame(label="Questions and Answers", wrap=True)

    run_button.click(
        fn=run_and_submit_all,
        outputs=[status_output, results_table]
    )

# --- Main ---
if __name__ == "__main__":
    logger.info("\n" + "-"*30 + " App Starting " + "-"*30)
    logger.info(f"SPACE_ID: {SPACE_ID}")
    logger.info("Launching Gradio Interface...")
    demo.launch(debug=True, share=False)