🏗️ Building on HF

Sergio Paniego PRO

sergiopaniego

250 213 138

wlau2023's profile picture

hiennguyen9874's profile picture

IlyasMoutawwakil's profile picture

https://sergiopaniego.github.io/

sergiopaniego
sergiopaniego
sergio-paniego-blanco

AI & ML interests

None yet

Recent Activity

updated a dataset about 3 hours ago

agents-course/certificates

updated a dataset about 24 hours ago

agents-course/final-certificates

updated a dataset about 24 hours ago

agents-course/course-certificates-of-excellence

View all activity

Organizations

sergiopaniego 's collections 15

MOPD with TRL: experiment

Experiment from the blog huggingface.co/blog/sergiopaniego/mopd-trl: Open-MOPD models and data, the 60-step checkpoint and its Trackio dashboard

BytedTsinghua-SIA/Open-MOPD-SmolLM3-3B-MixSFT

Text Generation • 3B • Updated Aug 18 • 982 • 1
BytedTsinghua-SIA/Open-MOPD-SmolLM3-3B-RL-Math

Text Generation • 3B • Updated Aug 18 • 380 • 1
BytedTsinghua-SIA/Open-MOPD-SmolLM3-3B-RL-Code

Text Generation • 3B • Updated Aug 18 • 383 • 1
BytedTsinghua-SIA/Open-MOPD-SmolLM3-3B-RL-IF

Text Generation • 3B • Updated Aug 18 • 309 • 1

Training Agents, Class 3: Reinforcement Learning

Training Agents Class 3 (GRPO): the training scripts, the Trackio dashboard and the task data.

sergiopaniego/grpo-youtube-livestream-3-scripts

Reinforcement Learning • Updated Jul 27 • 2
Running

Agents

1

Trackio Training Agents 3

🎯

1

Display an interactive tracking dashboard
google-research-datasets/mbpp

Viewer • Updated Jan 4, 2024 • 1.4k • 266k • 275
trl-lib/tldr

Viewer • Updated Jan 8, 2025 • 130k • 2.38k • 32

Training Agents · Class 4: RL Environments

Training coding agents inside RL environments with TRL and OpenEnv. Two demos, one per way of connecting an agent to a trainer.

sergiopaniego/rl-envs-youtube-livestream-4-scripts

Updated Sep 10
Running on CPU Upgrade

RL

Coding Environment Server

💻

Run Python code in an interactive coding environment
Running

Agents

Trackio Training Agents 4

🎯

Show live tracking data in a visual interface
sergiopaniego/qwen3-1.7b-mbpp-grpo

Text Generation • 2B • Updated Sep 10 • 29

Bringing Autonomous Driving RL to OpenEnv and TRL resources

Blog: https://huggingface.co/blog/sergiopaniego/bringing-carla-to-openenv-trl/

Runtime error

RL

CARLA Environment Server

🚗

Control a Carla driving simulation with custom actions
Runtime error

RL

CARLA Environment Server

🚗

Control a CARLA driving simulator with custom actions
Sleeping

Agents

Carla Grpo Trolley

🚀

Visualize your program’s I/O activity in real time
sergiopaniego/Qwen3-0.6B-carla-trolley-escape

0.8B • Updated Feb 26 • 4

Amazing design resources

Running

140

HFBA

🤗

140

Hugging Face brand assets — Huggies, logos, and graphics
Running

15

HF Thumbnail Crafter

🎨

15

Create custom thumbnails for your videos

GUI Grounding datasets

bevaya/ScreenSpot

Viewer • Updated Apr 10, 2024 • 1.27k • 3.26k • 52
OS-Copilot/OS-Atlas-data

Updated Dec 4, 2024 • 14.6k • 49

👁 Vision comparison ftw

Spaces to compare vision models — there’s no single best model, only the best one for your specific use case.

Running

Agents

43

comparevlms

🏃

43

Compare Vision Language Models
Running on Zero

Agents

68

OCR Time Machine

📚

68

Extract text from images and XML files using OCR models
Sleeping

Agents

26

Compare Docvqa Models

🦀

26

Compare different visual question answering
Running on CPU Upgrade

Agents

24

Compare Clip Siglip

🏃

24

Compare strong zero-shot image classification models

Vision Language Models: 2025 Update

This collection includes all the models, datasets and Spaces mentioned in the blog Vision Language Models: 2025 Update

Qwen/Qwen2.5-Omni-7B

Any-to-Any • 11B • Updated Apr 30, 2025 • 277k • 1.97k
Running

Agents

Featured

376

Qwen2.5 Omni 7B Demo

🏆

376

Chat with AI using text, voice, images, and video
Qwen2.5-Omni Technical Report

Paper • 2503.20215 • Published Mar 26, 2025 • 173
openbmb/MiniCPM-o-2_6

Any-to-Any • 9B • Updated Aug 18 • 40.8k • 1.33k

🎤 Talks

Slides from my talks

Running

Building AI in the open (UC3M)

🤗

Watch a talk on building AI in the open
Running

RL

19

Train the agent slides

🤖

19

Watch a talk on modern agent training techniques
Running

RL

Train the agent slides (PyTorch Madrid)

🤖

Watch a talk on modern agent training techniques
Running

11

Multi-harness RL slides

🎛

11

Watch a tutorial on training coding agents with RL environments

Lisbon AI 2026 · Training through a harness

An off-the-shelf coding agent, left untouched, trained through its own harness with TRL and OpenEnv.

sergiopaniego/opencode-rollout-trace

Updated Sep 21 • 1.15k
sergiopaniego/Qwen3-8B-opencode-deepcoder-grpo

Text Generation • 8B • Updated Jul 28 • 32 • 1
agentica-org/DeepCoder-Preview-Dataset

Viewer • Updated Apr 9, 2025 • 25k • 3.65k • 119
Sleeping

Agents

Opencode Hf Sandbox

🎯

Show tracking data visualization

Training Agents · Class 2: Distillation

Training Agents Class 2 (distillation): script, pi-mono-chat dataset, distilled Qwen3 checkpoints, and the 2 Trackio dashboards.

sergiopaniego/pi-mono-youtube-livestream-2-scripts

Updated Jul 6 • 2
sergiopaniego/pi-mono-chat

Viewer • Updated Jul 2 • 886 • 37
sergiopaniego/qwen3-0.6b-pimono-logit-kd-lr1e5

Text Generation • 0.6B • Updated Jul 2 • 43
sergiopaniego/qwen3-0.6b-pimono-logit-kd-lr2e5

Text Generation • 0.6B • Updated Jul 2 • 41

📝 Research & Long-Form Blog Posts

In-depth technical articles and research pieces published by Hugging Face

Running

4.13k

The Ultra-Scale Playbook

🌌

4.13k

The ultimate guide to training LLM on large GPU Clusters
Running on CPU Upgrade

Featured

3.36k

The Smol Training Playbook

📚

3.36k

The secrets to building world-class LLMs
Running

371

Evaluation Guidebook

📝

371

Explore LLM benchmark score trends over time
Running

236

FineVision: Open Data is All You Need

📝

236

A new open-source dataset for training VLMs

Vision reasoning datasets

deepcs233/Visual-CoT

Preview • Updated Mar 11, 2025 • 4.81k • 64
lmms-lab/multimodal-open-r1-8k-verified

Viewer • Updated Jan 27, 2025 • 7.69k • 685 • 76
leonardPKU/GEOQA_R1V_Train_8K

Viewer • Updated Feb 11, 2025 • 8.03k • 379 • 14
leonardPKU/clevr_cogen_a_train

Viewer • Updated Feb 2, 2025 • 70k • 2.79k • 41

My vision Spaces

Vision Spaces created by me

Running on Zero

Agents

Featured

127

VLM Object Understanding

🦀

127

Explore object detection, visual grounding, keypoint Detecti
Runtime error

Agents

4

VQA Autonomous Driving SmolVLM2

🌖

4

Visual Question Answering - Autonomous Driving - SmolVLM2

😎 Awesome vision Spaces

Spaces where I've collaborated or that I consider unique!

Running

Agents

43

comparevlms

🏃

43

Compare Vision Language Models
Runtime error

Agents

4

Gemma3 License Plate Detection

📈

4

Gemma 3 for license plate detection
Running on Zero

Agents

Featured

144

Gemma 3n E4B It

⚡

144

Chat with an AI that understands text, images, audio, and video
Running on Zero

Agents

Featured

46

Moondream3

🏢

46

Image and video tasks with moondream3.

MOPD with TRL: experiment

Experiment from the blog huggingface.co/blog/sergiopaniego/mopd-trl: Open-MOPD models and data, the 60-step checkpoint and its Trackio dashboard

BytedTsinghua-SIA/Open-MOPD-SmolLM3-3B-MixSFT

Text Generation • 3B • Updated Aug 18 • 982 • 1
BytedTsinghua-SIA/Open-MOPD-SmolLM3-3B-RL-Math

Text Generation • 3B • Updated Aug 18 • 380 • 1
BytedTsinghua-SIA/Open-MOPD-SmolLM3-3B-RL-Code

Text Generation • 3B • Updated Aug 18 • 383 • 1
BytedTsinghua-SIA/Open-MOPD-SmolLM3-3B-RL-IF

Text Generation • 3B • Updated Aug 18 • 309 • 1

🎤 Talks

Slides from my talks

Running

Building AI in the open (UC3M)

🤗

Watch a talk on building AI in the open
Running

RL

19

Train the agent slides

🤖

19

Watch a talk on modern agent training techniques
Running

RL

Train the agent slides (PyTorch Madrid)

🤖

Watch a talk on modern agent training techniques
Running

11

Multi-harness RL slides

🎛

11

Watch a tutorial on training coding agents with RL environments

Training Agents, Class 3: Reinforcement Learning

Training Agents Class 3 (GRPO): the training scripts, the Trackio dashboard and the task data.

sergiopaniego/grpo-youtube-livestream-3-scripts

Reinforcement Learning • Updated Jul 27 • 2
Running

Agents

1

Trackio Training Agents 3

🎯

1

Display an interactive tracking dashboard
google-research-datasets/mbpp

Viewer • Updated Jan 4, 2024 • 1.4k • 266k • 275
trl-lib/tldr

Viewer • Updated Jan 8, 2025 • 130k • 2.38k • 32

Lisbon AI 2026 · Training through a harness

An off-the-shelf coding agent, left untouched, trained through its own harness with TRL and OpenEnv.

sergiopaniego/opencode-rollout-trace

Updated Sep 21 • 1.15k
sergiopaniego/Qwen3-8B-opencode-deepcoder-grpo

Text Generation • 8B • Updated Jul 28 • 32 • 1
agentica-org/DeepCoder-Preview-Dataset

Viewer • Updated Apr 9, 2025 • 25k • 3.65k • 119
Sleeping

Agents

Opencode Hf Sandbox

🎯

Show tracking data visualization

Training Agents · Class 4: RL Environments

Training coding agents inside RL environments with TRL and OpenEnv. Two demos, one per way of connecting an agent to a trainer.

sergiopaniego/rl-envs-youtube-livestream-4-scripts

Updated Sep 10
Running on CPU Upgrade

RL

Coding Environment Server

💻

Run Python code in an interactive coding environment
Running

Agents

Trackio Training Agents 4

🎯

Show live tracking data in a visual interface
sergiopaniego/qwen3-1.7b-mbpp-grpo

Text Generation • 2B • Updated Sep 10 • 29

Training Agents · Class 2: Distillation

Training Agents Class 2 (distillation): script, pi-mono-chat dataset, distilled Qwen3 checkpoints, and the 2 Trackio dashboards.

sergiopaniego/pi-mono-youtube-livestream-2-scripts

Updated Jul 6 • 2
sergiopaniego/pi-mono-chat

Viewer • Updated Jul 2 • 886 • 37
sergiopaniego/qwen3-0.6b-pimono-logit-kd-lr1e5

Text Generation • 0.6B • Updated Jul 2 • 43
sergiopaniego/qwen3-0.6b-pimono-logit-kd-lr2e5

Text Generation • 0.6B • Updated Jul 2 • 41

Bringing Autonomous Driving RL to OpenEnv and TRL resources

Blog: https://huggingface.co/blog/sergiopaniego/bringing-carla-to-openenv-trl/

Runtime error

RL

CARLA Environment Server

🚗

Control a Carla driving simulation with custom actions
Runtime error

RL

CARLA Environment Server

🚗

Control a CARLA driving simulator with custom actions
Sleeping

Agents

Carla Grpo Trolley

🚀

Visualize your program’s I/O activity in real time
sergiopaniego/Qwen3-0.6B-carla-trolley-escape

0.8B • Updated Feb 26 • 4

📝 Research & Long-Form Blog Posts

In-depth technical articles and research pieces published by Hugging Face

Running

4.13k

The Ultra-Scale Playbook

🌌

4.13k

The ultimate guide to training LLM on large GPU Clusters
Running on CPU Upgrade

Featured

3.36k

The Smol Training Playbook

📚

3.36k

The secrets to building world-class LLMs
Running

371

Evaluation Guidebook

📝

371

Explore LLM benchmark score trends over time
Running

236

FineVision: Open Data is All You Need

📝

236

A new open-source dataset for training VLMs

Amazing design resources

Running

140

HFBA

🤗

140

Hugging Face brand assets — Huggies, logos, and graphics
Running

15

HF Thumbnail Crafter

🎨

15

Create custom thumbnails for your videos

Vision reasoning datasets

deepcs233/Visual-CoT

Preview • Updated Mar 11, 2025 • 4.81k • 64
lmms-lab/multimodal-open-r1-8k-verified

Viewer • Updated Jan 27, 2025 • 7.69k • 685 • 76
leonardPKU/GEOQA_R1V_Train_8K

Viewer • Updated Feb 11, 2025 • 8.03k • 379 • 14
leonardPKU/clevr_cogen_a_train

Viewer • Updated Feb 2, 2025 • 70k • 2.79k • 41

GUI Grounding datasets

bevaya/ScreenSpot

Viewer • Updated Apr 10, 2024 • 1.27k • 3.26k • 52
OS-Copilot/OS-Atlas-data

Updated Dec 4, 2024 • 14.6k • 49

My vision Spaces

Vision Spaces created by me

Running on Zero

Agents

Featured

127

VLM Object Understanding

🦀

127

Explore object detection, visual grounding, keypoint Detecti
Runtime error

Agents

4

VQA Autonomous Driving SmolVLM2

🌖

4

Visual Question Answering - Autonomous Driving - SmolVLM2

👁 Vision comparison ftw

Spaces to compare vision models — there’s no single best model, only the best one for your specific use case.

Running

Agents

43

comparevlms

🏃

43

Compare Vision Language Models
Running on Zero

Agents

68

OCR Time Machine

📚

68

Extract text from images and XML files using OCR models
Sleeping

Agents

26

Compare Docvqa Models

🦀

26

Compare different visual question answering
Running on CPU Upgrade

Agents

24

Compare Clip Siglip

🏃

24

Compare strong zero-shot image classification models

😎 Awesome vision Spaces

Spaces where I've collaborated or that I consider unique!

Running

Agents

43

comparevlms

🏃

43

Compare Vision Language Models
Runtime error

Agents

4

Gemma3 License Plate Detection

📈

4

Gemma 3 for license plate detection
Running on Zero

Agents

Featured

144

Gemma 3n E4B It

⚡

144

Chat with an AI that understands text, images, audio, and video
Running on Zero

Agents

Featured

46

Moondream3

🏢

46

Image and video tasks with moondream3.

Vision Language Models: 2025 Update

This collection includes all the models, datasets and Spaces mentioned in the blog Vision Language Models: 2025 Update

Qwen/Qwen2.5-Omni-7B

Any-to-Any • 11B • Updated Apr 30, 2025 • 277k • 1.97k
Running

Agents

Featured

376

Qwen2.5 Omni 7B Demo

🏆

376

Chat with AI using text, voice, images, and video
Qwen2.5-Omni Technical Report

Paper • 2503.20215 • Published Mar 26, 2025 • 173
openbmb/MiniCPM-o-2_6

Any-to-Any • 9B • Updated Aug 18 • 40.8k • 1.33k

Sergio Paniego PRO

AI & ML interests

Recent Activity

Organizations

sergiopaniego 's collections 15

Trackio Training Agents 3

Coding Environment Server

Trackio Training Agents 4

CARLA Environment Server

CARLA Environment Server

Carla Grpo Trolley

HFBA

HF Thumbnail Crafter

comparevlms

OCR Time Machine

Compare Docvqa Models

Compare Clip Siglip

Qwen2.5 Omni 7B Demo

Building AI in the open (UC3M)

Train the agent slides

Train the agent slides (PyTorch Madrid)

Multi-harness RL slides

Opencode Hf Sandbox

The Ultra-Scale Playbook

The Smol Training Playbook

Evaluation Guidebook

FineVision: Open Data is All You Need

VLM Object Understanding

VQA Autonomous Driving SmolVLM2

comparevlms

Gemma3 License Plate Detection

Gemma 3n E4B It

Moondream3

Building AI in the open (UC3M)

Train the agent slides

Train the agent slides (PyTorch Madrid)

Multi-harness RL slides

Trackio Training Agents 3

Opencode Hf Sandbox

Coding Environment Server

Trackio Training Agents 4

CARLA Environment Server

CARLA Environment Server

Carla Grpo Trolley

The Ultra-Scale Playbook

The Smol Training Playbook

Evaluation Guidebook

FineVision: Open Data is All You Need

HFBA

HF Thumbnail Crafter

VLM Object Understanding

VQA Autonomous Driving SmolVLM2

comparevlms

OCR Time Machine

Compare Docvqa Models

Compare Clip Siglip

comparevlms

Gemma3 License Plate Detection

Gemma 3n E4B It

Moondream3

Qwen2.5 Omni 7B Demo