model
string | evaluation_date
string | task_id
string | agent_type
string | difficulty
string | prompt
string | success
bool | tool_called
bool | correct_tool
bool | final_answer_called
bool | response_correct
bool | tools_used
list | steps
int64 | response
string | error
null | trace_id
string | execution_time_ms
float64 | total_tokens
int64 | cost_usd
float64 | enhanced_trace_info
string |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
tool_weather_single
|
tool
|
easy
|
What's the weather in Paris, France?
| true
| true
| true
| true
| true
|
[
"get_weather"
] | 3
|
The weather in Paris, France is 20°C and partly cloudy.
| null |
0x1ae60ae381ef8170bbb0c23a847ad5ca
| 16,815.8186
| 10,943
| 0.000738
|
{"trace_id": "0x1ae60ae381ef8170bbb0c23a847ad5ca", "total_tokens": 10943, "duration_ms": 16815.8186, "cost_usd": 0.0007375, "span_count": 11}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
tool_time_single
|
tool
|
easy
|
What time is it in UTC?
| true
| true
| true
| true
| true
|
[
"get_current_time"
] | 3
|
The current time in UTC is 2025-12-10 11:18:11.
| null |
0x68a04058bfd6db538bab9fdfb153fb84
| 6,611.2786
| 11,000
| 0.000742
|
{"trace_id": "0x68a04058bfd6db538bab9fdfb153fb84", "total_tokens": 11000, "duration_ms": 6611.2786, "cost_usd": 0.0007423, "span_count": 11}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
tool_search_single
|
tool
|
easy
|
Search for information about Python programming language
| true
| true
| true
| true
| true
|
[
"web_search"
] | 3
|
Python is a high-level, general-purpose programming language known for its emphasis on code readability and simplicity. It was created by Guido van Rossum in the late 1980s and supports multiple programming paradigms, including procedural, object-oriented, and functional programming. Python is dynamically typed and garbage-collected, making it versatile for various applications. It is widely used in machine learning, web development, and as an introductory programming language. Python 3.0, released in 2008, introduced significant changes, and the language continues to evolve with annual updates and long-term support for major versions. Python consistently ranks among the top programming languages in popularity and is supported by a robust community and extensive libraries.
| null |
0xfbd25ed18d4f5eb9fe3985bbfcbc0640
| 99,076.5352
| 14,436
| 0.001004
|
{"trace_id": "0xfbd25ed18d4f5eb9fe3985bbfcbc0640", "total_tokens": 14436, "duration_ms": 99076.53520000001, "cost_usd": 0.0010042000000000002, "span_count": 11}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
tool_weather_compare
|
tool
|
medium
|
Compare the weather in Paris, France and London, UK. Which one is warmer?
| true
| true
| true
| true
| true
|
[
"get_weather",
"get_weather"
] | 3
|
Paris, France is warmer with a temperature of 20°C compared to London, UK at 15°C.
| null |
0x9bd499b07b91c284c85f4f47626ca2c7
| 50,835.3502
| 11,199
| 0.000759
|
{"trace_id": "0x9bd499b07b91c284c85f4f47626ca2c7", "total_tokens": 11199, "duration_ms": 50835.3502, "cost_usd": 0.0007589000000000001, "span_count": 9}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
tool_search_and_summarize
|
tool
|
medium
|
Search for the latest news about AI and tell me what you find.
| true
| true
| true
| true
| true
|
[
"web_search"
] | 3
|
Here are some of the latest news and updates about AI:
1. **New AI Scenarios of the Future**: Insights into the future of AI and how to thrive in different scenarios. [Read more](https://www.bing.com/aclick?ld=e8rgeAOEtP45XFSkDd3mb5ATVUCUzXOkcp8_K1NbrJ0PMSdEYAlXnZGt-cZF97VcNlBV7O5WfyfCc-bCyR3ilSkpN0PLZPRIaOxPlof0dlmM4S8AeSKb6BWIvd34TYTGTWGG5CIIGguZ98Jfrsm4b_g10eZ849FW6vdmUcZPUjPQJdfNYJOCTd_KGsO3m6jgSxBodK9vhTXyqa0p_sv_sX5NAwODM&u=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&rlid=f9d59ec6427e1fec5b00f7ff4ba146bb)
2. **Google News - Artificial Intelligence**: Comprehensive coverage of AI-related topics, including articles, videos, and more. [Read more](https://news.google.com/topics/CAAqJAgKIh5DQkFTRUFvSEwyMHZNRzFyZWhJRlpXNHRSMElvQUFQAQ)
3. **TechCrunch - AI News**: Updates on artificial intelligence and machine learning technologies, companies, and ethical issues. [Read more](https://techcrunch.com/category/artificial-intelligence/)
4. **Reuters - AI News**: Latest developments in AI, including breakthroughs, trends, regulations, and global impact. [Read more](https://www.reuters.com/technology/artificial-intelligence/)
5. **ScienceDaily - Artificial Intelligence News**: Coverage of AI advancements, including futuristic robots and computer models of human intelligence. [Read more](https://www.sciencedaily.com/news/computers_math/artificial_intelligence/)
6. **AI News - Latest Headlines**: Updates on AI, machine learning, deep learning, and emerging technologies. [Read more](https://www.artificialintelligence-news.com/)
7. **Engadget - AI News**: News and reviews on AI software, hardware, and research. [Read more](https://www.engadget.com/ai/)
8. **Geeky Gadgets - Latest AI News**: Roundup of recent AI news, including updates on OpenAI's ChatGPT-5 and other major AI developments. [Read more](https://www.geeky-gadgets.com/latest-ai-news/)
9. **OWDT - AI News**: Insights and advancements in artificial intelligence. [Read more](https://owdt.com/tag/ai-news/)
10. **The Hacker News - AI Adoption**: News, reports, and analysis on AI adoption and industry insights. [Read more](https://thehackernews.com/search/label/AI+Adoption)
| null |
0xc20e2de599fe6fb50583d52eb03108a6
| 58,361.5433
| 17,864
| 0.001345
|
{"trace_id": "0xc20e2de599fe6fb50583d52eb03108a6", "total_tokens": 17864, "duration_ms": 58361.543300000005, "cost_usd": 0.0013445000000000002, "span_count": 11}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
tool_weather_time_combined
|
tool
|
hard
|
What's the current time in UTC and what's the weather in Tokyo, Japan?
| true
| true
| true
| true
| true
|
[
"get_current_time",
"get_weather"
] | 3
|
The current time in UTC is 2025-12-10 11:19:33, and the weather in Tokyo, Japan is 18°C and clear.
| null |
0xe48d22cb2eafdc47aced771ff2b4153f
| 10,526.6344
| 11,292
| 0.000768
|
{"trace_id": "0xe48d22cb2eafdc47aced771ff2b4153f", "total_tokens": 11292, "duration_ms": 10526.6344, "cost_usd": 0.0007684, "span_count": 9}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
shared_basic_weather
|
tool
|
easy
|
What's the weather like in Sydney, Australia?
| true
| true
| true
| true
| true
|
[
"get_weather"
] | 3
|
The weather in Sydney, Australia is currently 22°C and windy.
| null |
0xec8a5ca03be93b2bce9cade07400ab56
| 7,389.4551
| 10,945
| 0.000738
|
{"trace_id": "0xec8a5ca03be93b2bce9cade07400ab56", "total_tokens": 10945, "duration_ms": 7389.4551, "cost_usd": 0.0007377, "span_count": 11}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
shared_basic_search
|
tool
|
easy
|
Search for information about machine learning
| true
| true
| true
| true
| true
|
[
"web_search"
] | 3
|
Machine learning (ML) is a field of study in artificial intelligence focused on developing statistical algorithms that can learn from data and generalize to unseen data, enabling tasks to be performed without explicit instructions. It encompasses various techniques such as deep learning, supervised learning, and unsupervised learning, and finds applications in fields like natural language processing, computer vision, and predictive analytics. ML is built on foundations of statistics and mathematical optimization, and it is closely related to data mining and exploratory data analysis.
| null |
0x64dea839012968c5ff9f8ea0b548d53
| 24,735.5223
| 15,358
| 0.001131
|
{"trace_id": "0x64dea839012968c5ff9f8ea0b548d53", "total_tokens": 15358, "duration_ms": 24735.5223, "cost_usd": 0.0011306, "span_count": 11}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
code_calculator_single
|
code
|
easy
|
What is 234 multiplied by 67?
| true
| true
| false
| true
| true
|
[
"python_interpreter",
"python_interpreter"
] | 3
|
15678
| null |
0xd525c0224bfdfd5f12e6a49e28b82f5a
| 12,700.105
| 14,991
| 0.001017
|
{"trace_id": "0xd525c0224bfdfd5f12e6a49e28b82f5a", "total_tokens": 14991, "duration_ms": 12700.105, "cost_usd": 0.0010174000000000001, "span_count": 11}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
code_calculator_complex
|
code
|
medium
|
Calculate (450 + 230) * 3, then divide the result by 10
| true
| true
| false
| true
| true
|
[
"python_interpreter",
"python_interpreter",
"python_interpreter"
] | 4
|
Error calculating: invalid syntax (<string>, line 1)
| null |
0x1b5ce874ea62fddfa15faff93e820f3
| 24,307.1175
| 13,842
| 0.001183
|
{"trace_id": "0x1b5ce874ea62fddfa15faff93e820f3", "total_tokens": 13842, "duration_ms": 24307.1175, "cost_usd": 0.0011834, "span_count": 15}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
code_weather_with_calc
|
code
|
hard
|
Get the weather in Paris and if the temperature is above 15°C, calculate 15 * 2
| true
| true
| true
| true
| true
|
[
"python_interpreter",
"python_interpreter",
"python_interpreter"
] | 4
|
30
| null |
0xb07f6d27f362c3d1d759e3ec071e5016
| 20,112.9708
| 23,571
| 0.001609
|
{"trace_id": "0xb07f6d27f362c3d1d759e3ec071e5016", "total_tokens": 23571, "duration_ms": 20112.9708, "cost_usd": 0.0016088, "span_count": 15}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
code_search_calculate
|
code
|
hard
|
Search for the population of Paris, then if you find it's around 2 million, calculate what 2 million divided by 365 is
| true
| true
| true
| true
| true
|
[
"python_interpreter",
"python_interpreter",
"python_interpreter"
] | 4
|
5479.45205479452
| null |
0x835cf90a4ca70c0c16ae1d4bb8c63c33
| 51,104.533
| 31,427
| 0.00214
|
{"trace_id": "0x835cf90a4ca70c0c16ae1d4bb8c63c33", "total_tokens": 31427, "duration_ms": 51104.533, "cost_usd": 0.0021402, "span_count": 15}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
code_list_processing
|
code
|
hard
|
Get weather for Paris, London, and Tokyo, then tell me which cities have temperature above 18°C
| true
| true
| false
| true
| true
|
[
"python_interpreter",
"python_interpreter",
"python_interpreter",
"python_interpreter",
"python_interpreter",
"python_interpreter"
] | 7
|
Paris
| null |
0x75a7f5b75e65eb488e63a1d30b5917f8
| 70,329.028
| 67,551
| 0.004732
|
{"trace_id": "0x75a7f5b75e65eb488e63a1d30b5917f8", "total_tokens": 67551, "duration_ms": 70329.028, "cost_usd": 0.0047317, "span_count": 23}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
shared_basic_weather
|
code
|
easy
|
What's the weather like in Sydney, Australia?
| true
| true
| false
| true
| true
|
[
"python_interpreter",
"python_interpreter"
] | 3
|
The weather in Sydney, Australia is 22°C and windy.
| null |
0xec8a5ca03be93b2bce9cade07400ab56
| 7,389.4551
| 10,945
| 0.000738
|
{"trace_id": "0xec8a5ca03be93b2bce9cade07400ab56", "total_tokens": 10945, "duration_ms": 7389.4551, "cost_usd": 0.0007377, "span_count": 11}
|
mistral/devstral-medium-latest
|
2025-12-10T11:21:50.508599
|
shared_basic_search
|
code
|
easy
|
Search for information about machine learning
| true
| true
| false
| true
| true
|
[
"python_interpreter",
"python_interpreter"
] | 3
|
### Machine Learning Overview
**Definition**:
Machine Learning (ML) is a subfield of artificial intelligence (AI) that focuses on developing algorithms capable of learning from data and making predictions or decisions without explicit programming. It involves using statistical algorithms to identify patterns in data and generalize to unseen data.
**Types of Machine Learning**:
1. **Supervised Learning**: Uses labeled data to train models for tasks like classification and regression.
2. **Unsupervised Learning**: Involves finding patterns in unlabeled data, often used for clustering and association.
3. **Reinforcement Learning**: Focuses on training models to make sequences of decisions by rewarding desired behaviors.
**Applications**:
Machine learning is used in various fields such as healthcare, finance, autonomous vehicles, and natural language processing. It provides the backbone for many modern AI systems, including forecasting models, autonomous vehicles, and generative AI tools like large language models (LLMs).
### Latest Trends in Machine Learning (2025)
1. **Edge AI**: More prominent in industry, facilitating a shift in how new applications are developed.
2. **Smaller but Smarter Models**: Focus on efficient and effective models.
3. **Multimodal Generative AI**: Integration of multiple data types for more comprehensive AI applications.
4. **Autonomous AI Agents**: AI systems that can operate independently.
5. **Regulation-Driven Responsible AI Frameworks**: Emphasis on ethical and responsible AI development.
6. **Automated Machine Learning**: Streamlining the process of applying machine learning to real-world problems.
These trends highlight the evolving landscape of machine learning, emphasizing not only technological advancements but also ethical considerations and practical applications.
| null |
0x64dea839012968c5ff9f8ea0b548d53
| 24,735.5223
| 15,358
| 0.001131
|
{"trace_id": "0x64dea839012968c5ff9f8ea0b548d53", "total_tokens": 15358, "duration_ms": 24735.5223, "cost_usd": 0.0011306, "span_count": 11}
|
Tiny Agents. Total Visibility.
SMOLTRACE Evaluation Results
This dataset contains evaluation results from a SMOLTRACE benchmark run.
Dataset Information
| Field | Value |
|---|---|
| Model | mistral/devstral-medium-latest |
| Run ID | f9a1c3c3-95cb-4f3b-9e7c-208927e79239 |
| Agent Type | both |
| Total Tests | 15 |
| Generated | 2025-12-10 11:22:02 UTC |
| Source Dataset | kshitijthakkar/smoltrace-tasks |
Schema
| Column | Type | Description |
|---|---|---|
model |
string | Model identifier |
evaluation_date |
string | ISO timestamp of evaluation |
task_id |
string | Unique test case identifier |
agent_type |
string | "tool" or "code" agent type |
difficulty |
string | Test difficulty level |
prompt |
string | Test prompt/question |
success |
bool | Whether the test passed |
tool_called |
bool | Whether a tool was invoked |
correct_tool |
bool | Whether the correct tool was used |
final_answer_called |
bool | Whether final_answer was called |
response_correct |
bool | Whether the response was correct |
tools_used |
string | Comma-separated list of tools used |
steps |
int | Number of agent steps taken |
response |
string | Agent's final response |
error |
string | Error message if failed |
trace_id |
string | OpenTelemetry trace ID |
execution_time_ms |
float | Execution time in milliseconds |
total_tokens |
int | Total tokens consumed |
cost_usd |
float | API cost in USD |
enhanced_trace_info |
string | JSON with detailed trace data |
Usage
from datasets import load_dataset
# Load the results dataset
ds = load_dataset("YOUR_USERNAME/smoltrace-results-TIMESTAMP")
# Filter successful tests
successful = ds.filter(lambda x: x['success'])
# Calculate success rate
success_rate = sum(1 for r in ds['train'] if r['success']) / len(ds['train']) * 100
print(f"Success Rate: {success_rate:.2f}%")
Related Datasets
This evaluation run also generated:
- Traces Dataset: Detailed OpenTelemetry execution traces
- Metrics Dataset: GPU utilization and environmental metrics
- Leaderboard: Aggregated metrics for model comparison
About SMOLTRACE
SMOLTRACE is a comprehensive benchmarking and evaluation framework for Smolagents - HuggingFace's lightweight agent library.
Key Features
- Automated agent evaluation with customizable test cases
- OpenTelemetry-based tracing for detailed execution insights
- GPU metrics collection (utilization, memory, temperature, power)
- CO2 emissions and power cost tracking
- Leaderboard aggregation and comparison
Quick Links
Installation
pip install smoltrace
Citation
If you use SMOLTRACE in your research, please cite:
@software{smoltrace,
title = {SMOLTRACE: Benchmarking Framework for Smolagents},
author = {Thakkar, Kshitij},
url = {https://github.com/Mandark-droid/SMOLTRACE},
year = {2025}
}
Generated by SMOLTRACE
- Downloads last month
- 18