Robot Learning: A Tutorial
Paper
•
2510.12403
•
Published
•
104
AI & ML interests
None defined yet.
Recent Activity
View all activity
fracapuano
authored
a
paper
26 days ago
Post
3057
🚀 New blog: Maintain the unmaintainable – 1M+ Python LOC, 400+ models
How do you stop a million-line library built by thousands of contributors from collapsing under its own weight?
At 🤗 Transformers, we do it with explicit software-engineering tenets, principles that make the codebase hackable at scale.
🔍 Inside the post:
– One Model, One File: readability first — you can still open a modeling file and see the full logic, top to bottom.
– Modular Transformers: visible inheritance that cuts maintenance cost by ~15× while keeping models readable.
– Config-Driven Performance: FlashAttention, tensor parallelism, and attention scheduling are config-level features, not rewrites.
Written with @lysandre ,@pcuenq and @yonigozlan , this is a deep dive into how Transformers stays fast, open, and maintainable.
Read it here → transformers-community/Transformers-tenets
How do you stop a million-line library built by thousands of contributors from collapsing under its own weight?
At 🤗 Transformers, we do it with explicit software-engineering tenets, principles that make the codebase hackable at scale.
🔍 Inside the post:
– One Model, One File: readability first — you can still open a modeling file and see the full logic, top to bottom.
– Modular Transformers: visible inheritance that cuts maintenance cost by ~15× while keeping models readable.
– Config-Driven Performance: FlashAttention, tensor parallelism, and attention scheduling are config-level features, not rewrites.
Written with @lysandre ,@pcuenq and @yonigozlan , this is a deep dive into how Transformers stays fast, open, and maintainable.
Read it here → transformers-community/Transformers-tenets
tfrere
updated
a
Space
about 2 months ago
fracapuano
authored
a
paper
4 months ago
fracapuano
authored
a
paper
5 months ago
clefourrier
posted
an
update
6 months ago
Post
1818
Always surprised that so few people actually read the FineTasks blog, on
✨how to select training evals with the highest signal✨
If you're serious about training models without wasting compute on shitty runs, you absolutely should read it!!
An high signal eval actually tells you precisely, during training, how wel & what your model is learning, allowing you to discard the bad runs/bad samplings/...!
The blog covers in depth prompt choice, metrics, dataset, across languages/capabilities, and my fave section is "which properties should evals have"👌
(to know on your use case how to select the best evals for you)
Blog: HuggingFaceFW/blogpost-fine-tasks
✨how to select training evals with the highest signal✨
If you're serious about training models without wasting compute on shitty runs, you absolutely should read it!!
An high signal eval actually tells you precisely, during training, how wel & what your model is learning, allowing you to discard the bad runs/bad samplings/...!
The blog covers in depth prompt choice, metrics, dataset, across languages/capabilities, and my fave section is "which properties should evals have"👌
(to know on your use case how to select the best evals for you)
Blog: HuggingFaceFW/blogpost-fine-tasks
-
- 2 replies
clefourrier
posted
an
update
8 months ago
Post
2643
Gemma3 family is out! Reading the tech report, and this section was really interesting to me from a methods/scientific fairness pov.
Instead of doing over-hyped comparisons, they clearly state that **results are reported in a setup which is advantageous to their models**.
(Which everybody does, but people usually don't say)
For a tech report, it makes a lot of sense to report model performance when used optimally!
On leaderboards on the other hand, comparison will be apples to apples, but in a potentially unoptimal way for a given model family (like some user interact sub-optimally with models)
Also contains a cool section (6) on training data memorization rate too! Important to see if your model will output the training data it has seen as such: always an issue for privacy/copyright/... but also very much for evaluation!
Because if your model knows its evals by heart, you're not testing for generalization.
Instead of doing over-hyped comparisons, they clearly state that **results are reported in a setup which is advantageous to their models**.
(Which everybody does, but people usually don't say)
For a tech report, it makes a lot of sense to report model performance when used optimally!
On leaderboards on the other hand, comparison will be apples to apples, but in a potentially unoptimal way for a given model family (like some user interact sub-optimally with models)
Also contains a cool section (6) on training data memorization rate too! Important to see if your model will output the training data it has seen as such: always an issue for privacy/copyright/... but also very much for evaluation!
Because if your model knows its evals by heart, you're not testing for generalization.
clefourrier
authored
a
paper
9 months ago
clefourrier
authored
a
paper
11 months ago
fracapuano
posted
an
update
12 months ago
Post
1597
Sharing what we have built over the course of the weekend at the
@llamameta
hackathon, by Cerebral Valley in London 🇬🇧 👇
@gabrycina @calebgcc and I competed with 200+ participants and 50+ teams for a 24-hrs sprint centered around hacking for impact! We focused on applications of robotics to those in need of assisted living, moving our focus to enable greater autonomy and accessibility of robotics in everyday life.
complete list of assets 👇
🤗 trained robotics policies
v1:
- fracapuano/moss-pills
- fracapuano/moss-cup
v2:
- fracapuano/meta-grasp
🤗 datasets
v1:
- fracapuano/pills
- fracapuano/cup
v2:
- fracapuano/cupim
You can find a live demo of our submission at: https://x.com/_fracapuano/status/1858102728691458554
If you want to know more about how we collected 100GB+ of data, trained multiple RL-policies using @lerobot and used Llama-3.2 models to handle user interactions and switch between tasks, go ahead and have a look! Also, don't be a stranger, and reach out 🦾
Our project is fully open-source, for the community (and ourselves, 👨🍳) to build! A huge thank you to @cadene for the help (and the robot 🤭) - truly feeling these hugs-vibes 🤗 , and to @thomwolf and @clem for sharing our work across
Little extra:
➡️ Our 🧠EEG waves🧠-based control of the 🦾robotic arm🦾
@gabrycina @calebgcc and I competed with 200+ participants and 50+ teams for a 24-hrs sprint centered around hacking for impact! We focused on applications of robotics to those in need of assisted living, moving our focus to enable greater autonomy and accessibility of robotics in everyday life.
complete list of assets 👇
🤗 trained robotics policies
v1:
- fracapuano/moss-pills
- fracapuano/moss-cup
v2:
- fracapuano/meta-grasp
🤗 datasets
v1:
- fracapuano/pills
- fracapuano/cup
v2:
- fracapuano/cupim
You can find a live demo of our submission at: https://x.com/_fracapuano/status/1858102728691458554
If you want to know more about how we collected 100GB+ of data, trained multiple RL-policies using @lerobot and used Llama-3.2 models to handle user interactions and switch between tasks, go ahead and have a look! Also, don't be a stranger, and reach out 🦾
Our project is fully open-source, for the community (and ourselves, 👨🍳) to build! A huge thank you to @cadene for the help (and the robot 🤭) - truly feeling these hugs-vibes 🤗 , and to @thomwolf and @clem for sharing our work across
Little extra:
➡️ Our 🧠EEG waves🧠-based control of the 🦾robotic arm🦾
fracapuano
posted
an
update
12 months ago
Post
724
✍️ the last few weeks has been very intense!
🔴 I have been out all weekends
🔴 Participated in 4 hackathons in a row (2 more to come!)
🔴 Even threw a big hackathon myself!
Nonetheless, I am in school again 🏫, which meant... ✨homework✨
➡️ Head out to here https://x.com/_fracapuano/status/1856415612202799243 to read more about how I used @mistralai models to help me with my assignments (not how you think I did hihi 😏)
➡️ Check outhttps://huggingface.co/spaces/fracapuano/texstral if you want to use the tool yourself!
🔴 I have been out all weekends
🔴 Participated in 4 hackathons in a row (2 more to come!)
🔴 Even threw a big hackathon myself!
Nonetheless, I am in school again 🏫, which meant... ✨homework✨
➡️ Head out to here https://x.com/_fracapuano/status/1856415612202799243 to read more about how I used @mistralai models to help me with my assignments (not how you think I did hihi 😏)
➡️ Check outhttps://huggingface.co/spaces/fracapuano/texstral if you want to use the tool yourself!
clefourrier
authored
2
papers
over 1 year ago
clefourrier
posted
an
update
over 1 year ago
Post
6164
In a basic chatbots, errors are annoyances. In medical LLMs, errors can have life-threatening consequences 🩸
It's therefore vital to benchmark/follow advances in medical LLMs before even thinking about deployment.
This is why a small research team introduced a medical LLM leaderboard, to get reproducible and comparable results between LLMs, and allow everyone to follow advances in the field.
openlifescienceai/open_medical_llm_leaderboard
Congrats to @aaditya and @pminervini !
Learn more in the blog: https://huggingface.co/blog/leaderboard-medicalllm
It's therefore vital to benchmark/follow advances in medical LLMs before even thinking about deployment.
This is why a small research team introduced a medical LLM leaderboard, to get reproducible and comparable results between LLMs, and allow everyone to follow advances in the field.
openlifescienceai/open_medical_llm_leaderboard
Congrats to @aaditya and @pminervini !
Learn more in the blog: https://huggingface.co/blog/leaderboard-medicalllm
clefourrier
posted
an
update
over 1 year ago
Post
4795
Contamination free code evaluations with LiveCodeBench! 🖥️
LiveCodeBench is a new leaderboard, which contains:
- complete code evaluations (on code generation, self repair, code execution, tests)
- my favorite feature: problem selection by publication date 📅
This feature means that you can get model scores averaged only on new problems out of the training data. This means... contamination free code evals! 🚀
Check it out!
Blog: https://huggingface.co/blog/leaderboard-livecodebench
Leaderboard: livecodebench/leaderboard
Congrats to @StringChaos @minimario @xu3kev @kingh0730 and @FanjiaYan for the super cool leaderboard!
LiveCodeBench is a new leaderboard, which contains:
- complete code evaluations (on code generation, self repair, code execution, tests)
- my favorite feature: problem selection by publication date 📅
This feature means that you can get model scores averaged only on new problems out of the training data. This means... contamination free code evals! 🚀
Check it out!
Blog: https://huggingface.co/blog/leaderboard-livecodebench
Leaderboard: livecodebench/leaderboard
Congrats to @StringChaos @minimario @xu3kev @kingh0730 and @FanjiaYan for the super cool leaderboard!
clefourrier
posted
an
update
over 1 year ago
Post
2276
🆕 Evaluate your RL agents - who's best at Atari?🏆
The new RL leaderboard evaluates agents in 87 possible environments (from Atari 🎮 to motion control simulations🚶and more)!
When you submit your model, it's run and evaluated in real time - and the leaderboard displays small videos of the best model's run, which is super fun to watch! ✨
Kudos to @qgallouedec for creating and maintaining the leaderboard!
Let's find out which agent is the best at games! 🚀
open-rl-leaderboard/leaderboard
The new RL leaderboard evaluates agents in 87 possible environments (from Atari 🎮 to motion control simulations🚶and more)!
When you submit your model, it's run and evaluated in real time - and the leaderboard displays small videos of the best model's run, which is super fun to watch! ✨
Kudos to @qgallouedec for creating and maintaining the leaderboard!
Let's find out which agent is the best at games! 🚀
open-rl-leaderboard/leaderboard
clefourrier
posted
an
update
over 1 year ago
Post
2264
Fun fact about evaluation, part 2!
How much do scores change depending on prompt format choice?
Using different prompts (all present in the literature, from
10 points for a single model!
Keep in mind that we only changed the prompt, not the evaluation subsets, etc.
Again, this confirms that evaluation results reported without their details are basically bullshit.
Prompt format on the x axis, all these evals look at the logprob of either "choice A/choice B..." or "A/B...".
Incidentally, it also changes model rankings - so a "best" model might only be best on one type of prompt...
How much do scores change depending on prompt format choice?
Using different prompts (all present in the literature, from
Prompt question? to Question: prompt question?\nChoices: enumeration of all choices\nAnswer: ), we get a score range of...10 points for a single model!
Keep in mind that we only changed the prompt, not the evaluation subsets, etc.
Again, this confirms that evaluation results reported without their details are basically bullshit.
Prompt format on the x axis, all these evals look at the logprob of either "choice A/choice B..." or "A/B...".
Incidentally, it also changes model rankings - so a "best" model might only be best on one type of prompt...
Post
5520
🚀🚀 Exciting times for the document AI community!
We're thrilled to announce the release of some of the largest OCR datasets available to the public.
🔥 With over 26 million pages , 18 billion text tokens, and 6TB of data, these resources are a significant leap forward for document AI research.
Here's how to access these datasets quickly:
This enables you to stream them directly, integrating seamlessly with your projects using the Hugging Face datasets library. On the hub, you can find them here:
pixparse/pdfa-eng-wds
pixparse/idl-wds
For lean data loading, the new [chug](https://github.com/huggingface/chug) library offers a solution with pdf decoding:
We owe a huge thank you to Peter Wyatt, Kate Tasker, Rachel Taketa, Ali Furkan Biten, Ruben Tito, and their colleagues for their contributions. Their work putting these datasets together has been invaluable. 🤗
Looking Ahead:
We're on a mission to enhance document AI capabilities, and these datasets are just the beginning. With your engagement and innovation, we're confident in the community's ability to develop robust OCR solutions. We encourage you to explore these datasets, experiment with the code, and contribute to the collective progress in document AI.
For detailed information on usage and licensing, please refer to the dataset cards on the Hugging Face hub.
We're thrilled to announce the release of some of the largest OCR datasets available to the public.
🔥 With over 26 million pages , 18 billion text tokens, and 6TB of data, these resources are a significant leap forward for document AI research.
Here's how to access these datasets quickly:
from datasets import load_dataset
pdfa_dataset = load_dataset('pixparse/pdfa-eng-wds', streaming=True)
IDL_dataset = load_dataset('pixparse/idl-wds', streaming=True)This enables you to stream them directly, integrating seamlessly with your projects using the Hugging Face datasets library. On the hub, you can find them here:
pixparse/pdfa-eng-wds
pixparse/idl-wds
For lean data loading, the new [chug](https://github.com/huggingface/chug) library offers a solution with pdf decoding:
import chug
task_cfg = chug.DataTaskDocReadCfg(
page_sampling='all',
)
data_cfg = chug.DataCfg(
source='pixparse/pdfa-eng-wds',
split='train',
batch_size=None,
format='hfids',
num_workers=0,
)
data_loader = chug.create_loader(
data_cfg,
task_cfg,
)
sample = next(iter(data_loader))We owe a huge thank you to Peter Wyatt, Kate Tasker, Rachel Taketa, Ali Furkan Biten, Ruben Tito, and their colleagues for their contributions. Their work putting these datasets together has been invaluable. 🤗
Looking Ahead:
We're on a mission to enhance document AI capabilities, and these datasets are just the beginning. With your engagement and innovation, we're confident in the community's ability to develop robust OCR solutions. We encourage you to explore these datasets, experiment with the code, and contribute to the collective progress in document AI.
For detailed information on usage and licensing, please refer to the dataset cards on the Hugging Face hub.
- 4 replies
clefourrier
posted
an
update
over 1 year ago
Post
2394
Fun fact about evaluation!
Did you know that, if you evaluate the same model, with the same prompt formatting & the same fixed few-shot examples, only changing
♻️the order in which the few shot examples are added to the prompt ♻️
you get a difference of up to 3 points in evaluation score?
I did a small experiment using some MMLU subsets on the best performing 7B and lower pretrained models from the leaderboard.
I tried 8 different prompting methods (containing more or less information, such as just the question, or Question: question, or Question: question Choices: ..., see the x axis) that are commonly used in evaluation.
I then compared the results for all these methods, in 5-shot, during 2 runs. The *only difference* between the first and second run being that the samples used in few-shot are not introduced in the same order.
For example, run one would have been "A B C D E Current sample", vs, in run 2, "D C E A B Current sample".
All the other experiment parameters stayed exactly the same.
As you can see on the attached picture, you get a difference of up to 3 points between the 2 few-shot samples shuffling.
So, when just changing *the order of the few shot samples* can change your results by several points, what is the impact of all other "minimal" and unreported prompting changes?
-> Any kind of model score, provided without an evaluation script for reproducibility, is basically bullshit (or coms).
-> This is why we need reproducible evaluation in a fair and exactly similar setup, using evaluation suites such as
Did you know that, if you evaluate the same model, with the same prompt formatting & the same fixed few-shot examples, only changing
♻️the order in which the few shot examples are added to the prompt ♻️
you get a difference of up to 3 points in evaluation score?
I did a small experiment using some MMLU subsets on the best performing 7B and lower pretrained models from the leaderboard.
I tried 8 different prompting methods (containing more or less information, such as just the question, or Question: question, or Question: question Choices: ..., see the x axis) that are commonly used in evaluation.
I then compared the results for all these methods, in 5-shot, during 2 runs. The *only difference* between the first and second run being that the samples used in few-shot are not introduced in the same order.
For example, run one would have been "A B C D E Current sample", vs, in run 2, "D C E A B Current sample".
All the other experiment parameters stayed exactly the same.
As you can see on the attached picture, you get a difference of up to 3 points between the 2 few-shot samples shuffling.
So, when just changing *the order of the few shot samples* can change your results by several points, what is the impact of all other "minimal" and unreported prompting changes?
-> Any kind of model score, provided without an evaluation script for reproducibility, is basically bullshit (or coms).
-> This is why we need reproducible evaluation in a fair and exactly similar setup, using evaluation suites such as
lm_eval from the Harness, lighteval from HF, or the Open LLM Leaderboard.
-
- 4 replies
clefourrier
posted
an
update
over 1 year ago
Post
2040
Are you looking for the perfect leaderboard/arena for your use case? 👀
There's a new tool for this!
https://huggingface.co/spaces/leaderboards/LeaderboardFinder
Select your modality, language, task... then search! 🔍
Some categories of interest:
- does the leaderboard accept submissions?
- is the test set private or public?
- is it using an automatic metric, human evaluators, or llm as a judge?
The spaces list is build from space metadata, and reloaded every hour.
Enjoy!
There's a new tool for this!
https://huggingface.co/spaces/leaderboards/LeaderboardFinder
Select your modality, language, task... then search! 🔍
Some categories of interest:
- does the leaderboard accept submissions?
- is the test set private or public?
- is it using an automatic metric, human evaluators, or llm as a judge?
The spaces list is build from space metadata, and reloaded every hour.
Enjoy!