Gaia2 and ARE: Empowering the community to study agents
•
114
AI & ML interests
State-of-the-art Machine Learning for real-world robotics
Recent Activity
View all activity
Articles
aractingi
updated
a
dataset
about 11 hours ago
aractingi
updated
a
dataset
1 day ago
aractingi
updated
a
model
1 day ago
aractingi
updated
2
datasets
1 day ago
aractingi
published
a
model
4 days ago
fracapuano
authored
a
paper
8 days ago
aractingi
authored
a
paper
8 days ago
thomwolf
authored
a
paper
8 days ago
tfrere
updated
a
Space
9 days ago
fracapuano
updated
a
Space
9 days ago
thomwolf
updated
a
Space
9 days ago
fracapuano
published
a
Space
9 days ago
Convert smolvla with Pipeline
#11 opened about 1 month ago
by
AdilZtn
Post
2913
🚀 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
add lerobot tag
#3 opened 21 days ago
by
clem
add license and language tag
#2 opened 21 days ago
by
clem
add license and language tag
#2 opened 21 days ago
by
clem
add robotics tag
#1 opened 21 days ago
by
clem
pepijn223
updated
a
collection
22 days ago