File size: 16,123 Bytes
10e9b7d
eccf8e4
8fcb930
 
3c4371f
8fcb930
15b3a99
6492713
15b3a99
6dc5982
 
 
 
15b3a99
10e9b7d
613fdf5
3925f9b
613fdf5
 
 
 
 
d59f015
e80aab9
3db6293
e80aab9
24bb0c6
8e4d2c9
 
 
 
c835a09
a43a8a2
8b4bcba
 
8e4d2c9
 
24bb0c6
 
8b4bcba
24bb0c6
 
8e4d2c9
c835a09
 
8b4bcba
8e4d2c9
31243f4
d59f015
31243f4
613fdf5
8e4d2c9
613fdf5
 
31243f4
8e4d2c9
785687c
 
 
 
 
 
 
 
 
 
8e4d2c9
 
613fdf5
 
8e4d2c9
 
120a2b5
4802fca
8e4d2c9
78e3211
120a2b5
8e4d2c9
2418331
 
 
3c5c351
 
2418331
3c5c351
 
 
 
 
 
 
 
 
2418331
 
 
 
 
 
a43a8a2
2418331
 
 
8e4d2c9
 
 
 
 
 
 
613fdf5
8e4d2c9
613fdf5
03aa28e
613fdf5
8b4bcba
613fdf5
03aa28e
613fdf5
03aa28e
613fdf5
8b4bcba
 
03aa28e
 
613fdf5
8e4d2c9
 
 
 
613fdf5
 
8e4d2c9
 
 
613fdf5
8e4d2c9
1ec9653
8e4d2c9
 
1ec9653
8e4d2c9
 
 
 
24bb0c6
1ec9653
8e4d2c9
 
613fdf5
8e4d2c9
6492713
 
613fdf5
8e4d2c9
 
 
 
 
 
 
 
 
 
 
 
 
613fdf5
8e4d2c9
 
 
 
 
 
 
 
 
 
 
 
 
4021bf3
1ec9653
6492713
 
 
 
019c212
 
30b42bb
6492713
 
 
b90251f
31243f4
 
 
 
7d65c66
b177367
3c4371f
7e4a06b
1ca9f65
3c4371f
7e4a06b
3c4371f
7d65c66
3c4371f
7e4a06b
31243f4
 
e80aab9
b177367
31243f4
 
 
3c4371f
31243f4
b177367
36ed51a
c1fd3d2
3c4371f
7d65c66
31243f4
eccf8e4
31243f4
7d65c66
31243f4
 
3c4371f
 
31243f4
e80aab9
31243f4
 
3c4371f
 
7d65c66
3c4371f
7d65c66
31243f4
 
e80aab9
b177367
7d65c66
 
3c4371f
31243f4
 
 
 
 
 
 
7d65c66
 
 
31243f4
30b42bb
 
 
 
31243f4
 
3c4371f
31243f4
 
b177367
7d65c66
3c4371f
31243f4
e80aab9
7d65c66
31243f4
e80aab9
7d65c66
e80aab9
 
31243f4
e80aab9
 
3c4371f
 
 
e80aab9
 
31243f4
 
e80aab9
3c4371f
e80aab9
 
3c4371f
e80aab9
7d65c66
3c4371f
31243f4
7d65c66
31243f4
3c4371f
 
 
 
 
e80aab9
31243f4
 
 
 
7d65c66
31243f4
 
 
 
e80aab9
 
 
 
31243f4
0ee0419
e514fd7
 
 
81917a3
e514fd7
 
 
 
 
 
 
 
e80aab9
 
7e4a06b
e80aab9
31243f4
e80aab9
9088b99
7d65c66
 
e80aab9
31243f4
 
 
e80aab9
 
 
3c4371f
7d65c66
3c4371f
7d65c66
 
3c4371f
 
7d65c66
3c4371f
7d65c66
 
 
 
 
 
 
 
 
3c4371f
 
31243f4
3c4371f
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
import gradio as gr
import requests
import json
import torch
import pandas as pd
import inspect
import os
import gc

import torch, transformers
print(f"PyTorch: {torch.__version__}")
print(f"Transformers: {transformers.__version__}")

os.environ['CUDA_LAUNCH_BLOCKING'] = '1'

try:
    from smolagents import CodeAgent, DuckDuckGoSearchTool, TransformersModel
    SMOLAGENTS_AVAILABLE = True
except ImportError:
    print("Warning: smolagents not available, using fallback implementation")
    SMOLAGENTS_AVAILABLE = False

# (Keep Constants as is)
# --- Constants ---
DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"

CURRENT_PROMPT = """You are a helpful assistant that can use tools including web-search and writing Python code to answer questions.

Question: {question}

To answer this question:
1. If you need current information or facts you're unsure about, use the web_search tool
2. If you need to transcribe audio, import smolagents and instantiate 'transcriber = smolagents.Transcriber()'
3. Write Python code to solve the problem
4. Make sure your final answer is clear and direct

Available tools:
- web_search(query: str) β†’ str
- visit_webpage(url: str) β†’ str
- transcriber(audio_file_path: str) β†’ str
- python(code: str) β†’ Any
- final_answer(answer: str) β†’ None

Note: add "Wikipedia" to web searches in order to get expert results from Wikipedia.

Please provide a complete solution that ends with the correct final answer."""

# --- Basic Agent Definition ---
# ----- THIS IS WERE YOU CAN BUILD WHAT YOU WANT ------
class BasicAgent:
    """
    Basic agent using smolagents CodeAgent with DuckDuckGoSearchTool.
    """
    
    def __init__(self):
        print("BasicAgent initialized.")

        # Check GPU availability
        print(f"πŸ” GPU Check:")
        print(f"   - CUDA available: {torch.cuda.is_available()}")
        if torch.cuda.is_available():
            print(f"   - CUDA device count: {torch.cuda.device_count()}")
            print(f"   - Current device: {torch.cuda.current_device()}")
            print(f"   - Device name: {torch.cuda.get_device_name()}")
            print(f"   - Device memory: {torch.cuda.get_device_properties(0).total_memory / 1e9:.1f} GB")
        else:
            print("   - No CUDA devices found, will use CPU")

        if SMOLAGENTS_AVAILABLE:
            try:
                # Initialize the model
                print("πŸ€– Initializing TransformersModel...")
                self.model = TransformersModel(
                    model_id="Qwen/Qwen2.5-Coder-14B",
                    torch_dtype=torch.bfloat16,
                    device_map="auto",
                )

                if hasattr(self.model, 'tokenizer') and self.model.tokenizer is not None:
                    # Set left padding for better batching with causal models
                    self.model.tokenizer.padding_side = "left"
                    
                    # Set pad token properly - don't use eos_token as pad_token
                    if self.model.tokenizer.pad_token is None:
                        # Use a different token than eos for padding
                        if hasattr(self.model.tokenizer, 'unk_token') and self.model.tokenizer.unk_token is not None:
                            self.model.tokenizer.pad_token = self.model.tokenizer.unk_token
                        else:
                            # Add a new pad token
                            self.model.tokenizer.add_special_tokens({'pad_token': '[PAD]'})
                    
                    # Don't set pad_to_multiple_of for now as it may cause issues
                    print("βœ… Applied tokenizer padding fix")
                
                # If the model has a processor with tokenizer, fix that too
                if hasattr(self.model, 'processor') and hasattr(self.model.processor, 'tokenizer'):
                    self.model.processor.tokenizer.padding_side = "left"
                    if self.model.processor.tokenizer.pad_token is None:
                        self.model.processor.tokenizer.pad_token = self.model.processor.tokenizer.eos_token
                    self.model.processor.tokenizer.pad_to_multiple_of = 64
                    print("βœ… Applied processor tokenizer padding fix")


                # Verify where model actually loaded
                if hasattr(self.model, 'device'):
                    print(f"βœ… Model loaded on device: {self.model.device}")
                elif hasattr(self.model, 'model') and hasattr(self.model.model, 'device'):
                    print(f"βœ… Model loaded on device: {self.model.model.device}")
                else:
                    print("βœ… Model loaded (device info not directly accessible)")
                
                # Create CodeAgent with DuckDuckGoSearchTool and additional imports
                self.agent = CodeAgent(
                    tools=[],
                    model=self.model,
                    max_steps=24,
                    additional_authorized_imports=[
                        'math', 'statistics', 're',   # Basic computation
                        'requests', 'json',           # Web requests and JSON
                        'pandas', 'numpy', 'openpyxl',# Data analysis
                        'zipfile', 'os',              # File processing
                        'datetime', 'time',           # Date/time operations
                        'smolagents'
                    ],
                    add_base_tools=True,
                )

                self.tools_available = True
                print("βœ… Smolagents CodeAgent initialized with DuckDuckGoSearchTool")

            except Exception as e:
                print(f"⚠️ Error initializing smolagents: {e}")
                import traceback
                traceback.print_exc()
                self.tools_available = False
        else:
            self.tools_available = False

        if not self.tools_available:
            print("⚠️ Using fallback implementation without smolagents")

    def _run_smolagents(self, question):
        """Run question through smolagents CodeAgent with enhanced prompting."""
        try:
            # Use the global CURRENT_PROMPT variable
            formatted_question = CURRENT_PROMPT.format(question=question)

            print(f"πŸ”„ Processing question: {question}")
            print(f"πŸ”§ Available tools: {[tool.__class__.__name__ for tool in self.agent.tools]}")
            
            # Run the agent
            with torch.no_grad():
                result = self.agent.run(formatted_question)
            
            print(f"Raw result: {result}")

            # Clean up the result (remove any remaining prefixes)
            if isinstance(result, str):
                result = result.strip()
                # Remove common prefixes
                prefixes_to_remove = ["The answer is ", "Answer: ", "Final answer: "]
                for prefix in prefixes_to_remove:
                    if result.startswith(prefix):
                        result = result[len(prefix):].strip()

            return result

        except Exception as e:
            import traceback
            return f"Agent error: {e}\n{traceback.format_exc()}"

    def _fallback_implementation(self, question):
        """Fallback when smolagents is not available."""
        return f"Smolagents not available. Question received: {question}"

    def __call__(self, question):
        """Process a question using the smolagents CodeAgent or fallback."""
        if self.tools_available:
            return self._run_smolagents(question)
        else:
            return self._fallback_implementation(question)


def cleanup_memory():
    """Centralized memory cleanup function"""
    if torch.cuda.is_available():
        torch.cuda.synchronize()
        import time
        time.sleep(0.1)
        torch.cuda.empty_cache()
    gc.collect()


def run_and_submit_all( profile: gr.OAuthProfile | None):
    """
    Fetches all questions, runs the BasicAgent on them, submits all answers,
    and displays the results.
    """
    # --- Determine HF Space Runtime URL and Repo URL ---
    space_id = os.getenv("SPACE_ID") # Get the SPACE_ID for sending link to the code

    if profile:
        username= f"{profile.username}"
        print(f"User logged in: {username}")
    else:
        print("User not logged in.")
        return "Please Login to Hugging Face with the button.", None

    api_url = DEFAULT_API_URL
    questions_url = f"{api_url}/questions"
    submit_url = f"{api_url}/submit"

    # 1. Instantiate Agent ( modify this part to create your agent)
    try:
        agent = BasicAgent()
    except Exception as e:
        print(f"Error instantiating agent: {e}")
        return f"Error initializing agent: {e}", None
    # In the case of an app running as a hugging Face space, this link points toward your codebase ( usefull for others so please keep it public)
    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
    print(agent_code)

    # 2. Fetch Questions
    print(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:
             print("Fetched questions list is empty.")
             return "Fetched questions list is empty or invalid format.", None
        print(f"Fetched {len(questions_data)} questions.")
    except requests.exceptions.RequestException as e:
        print(f"Error fetching questions: {e}")
        return f"Error fetching questions: {e}", None
    except requests.exceptions.JSONDecodeError as e:
         print(f"Error decoding JSON response from questions endpoint: {e}")
         print(f"Response text: {response.text[:500]}")
         return f"Error decoding server response for questions: {e}", None
    except Exception as e:
        print(f"An unexpected error occurred fetching questions: {e}")
        return f"An unexpected error occurred fetching questions: {e}", None

    # 3. Run your Agent
    results_log = []
    answers_payload = []
    print(f"Running agent on {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:
            print(f"Skipping item with missing task_id or question: {item}")
            continue
        try:
            submitted_answer = agent(question_text)
            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:
            print(f"Error running agent on task {task_id}: {e}")
            results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": f"AGENT ERROR: {e}"})
        finally:
            cleanup_memory()

    if not answers_payload:
        print("Agent did not produce any answers to submit.")
        return "Agent did not produce any answers to submit.", pd.DataFrame(results_log)

    # 4. Prepare Submission 
    submission_data = {"username": username.strip(), "agent_code": agent_code, "answers": answers_payload}
    status_update = f"Agent finished. Submitting {len(answers_payload)} answers for user '{username}'..."
    print(status_update)

    # 5. Submit
    print(f"Submitting {len(answers_payload)} answers to: {submit_url}")
    try:
        response = requests.post(submit_url, json=submission_data, timeout=60)
        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.')}"
        )
        print("Submission successful.")
        results_df = pd.DataFrame(results_log)
        return final_status, results_df
    except requests.exceptions.HTTPError as e:
        error_detail = f"Server responded with status {e.response.status_code}."
        try:
            error_json = e.response.json()
            error_detail += f" Detail: {error_json.get('detail', e.response.text)}"
        except requests.exceptions.JSONDecodeError:
            error_detail += f" Response: {e.response.text[:500]}"
        status_message = f"Submission Failed: {error_detail}"
        print(status_message)
        results_df = pd.DataFrame(results_log)
        return status_message, results_df
    except requests.exceptions.Timeout:
        status_message = "Submission Failed: The request timed out."
        print(status_message)
        results_df = pd.DataFrame(results_log)
        return status_message, results_df
    except requests.exceptions.RequestException as e:
        status_message = f"Submission Failed: Network error - {e}"
        print(status_message)
        results_df = pd.DataFrame(results_log)
        return status_message, results_df
    except Exception as e:
        status_message = f"An unexpected error occurred during submission: {e}"
        print(status_message)
        results_df = pd.DataFrame(results_log)
        return status_message, results_df


# --- Build Gradio Interface using Blocks ---
with gr.Blocks() as demo:
    gr.Markdown("# Basic Agent Evaluation Runner")
    gr.Markdown(
        """
        **Instructions:**

        1.  Please clone this space, then modify the code to define your agent's logic, the tools, the necessary packages, etc ...
        2.  Log in to your Hugging Face account using the button below. This uses your HF username for submission.
        3.  Click 'Run Evaluation & Submit All Answers' to fetch questions, run your agent, submit answers, and see the score.

        ---
        **Disclaimers:**
        Once clicking on the "submit button, it can take quite some time ( this is the time for the agent to go through all the questions).
        This space provides a basic setup and is intentionally sub-optimal to encourage you to develop your own, more robust solution. For instance for the delay process of the submit button, a solution could be to cache the answers and submit in a seperate action or even to answer the questions in async.
        """
    )

    gr.LoginButton()

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

    status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
    # Removed max_rows=10 from DataFrame constructor
    results_table = gr.DataFrame(label="Questions and Agent Answers", wrap=True)

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

if __name__ == "__main__":
    print("\n" + "-"*30 + " App Starting " + "-"*30)
    # Check for SPACE_HOST and SPACE_ID at startup for information
    space_host_startup = os.getenv("SPACE_HOST")
    space_id_startup = os.getenv("SPACE_ID") # Get SPACE_ID at startup

    if space_host_startup:
        print(f"βœ… SPACE_HOST found: {space_host_startup}")
        print(f"   Runtime URL should be: https://{space_host_startup}.hf.space")
    else:
        print("ℹ️  SPACE_HOST environment variable not found (running locally?).")

    if space_id_startup: # Print repo URLs if SPACE_ID is found
        print(f"βœ… SPACE_ID found: {space_id_startup}")
        print(f"   Repo URL: https://huggingface.co/spaces/{space_id_startup}")
        print(f"   Repo Tree URL: https://huggingface.co/spaces/{space_id_startup}/tree/main")
    else:
        print("ℹ️  SPACE_ID environment variable not found (running locally?). Repo URL cannot be determined.")

    print("-"*(60 + len(" App Starting ")) + "\n")

    print("Launching Gradio Interface for Basic Agent Evaluation...")
    demo.launch(debug=True, share=False)