Introducing smolagents: simple agents that write actions in code.
โข
153
Aymeric Roucher
m-ric
AI & ML interests
Leading Agents at Hugging Face ๐ค
Recent Activity
updated
a Space
1 day ago
m-ric/get-travel-duration-tool
liked
a Space
3 days ago
reach-vb/2024-ai-timeline
updated
a dataset
4 days ago
huggingface/documentation-images
Articles
Organizations
m-ric's activity
updated
a
Space
1 day ago
liked
a
Space
3 days ago
updated
a
dataset
4 days ago
Upload _versions.yml
1
#32 opened 5 days ago
by
m-ric
updated
a
Space
4 days ago
upvoted
a
paper
5 days ago
liked
a
dataset
5 days ago
updated
a
dataset
5 days ago
liked
a
Space
7 days ago
liked
a
dataset
8 days ago
updated
a
dataset
8 days ago
liked
a
dataset
8 days ago
liked
a
model
9 days ago
updated
a
dataset
11 days ago
liked
a
Space
15 days ago
posted
an
update
15 days ago
Post
1898
After 6 years, BERT, the workhorse of encoder models, finally gets a replacement: ๐ช๐ฒ๐น๐ฐ๐ผ๐บ๐ฒ ๐ ๐ผ๐ฑ๐ฒ๐ฟ๐ป๐๐๐ฅ๐ง! ๐ค
We talk a lot about โจGenerative AIโจ, meaning "Decoder version of the Transformers architecture", but this is only one of the ways to build LLMs: encoder models, that turn a sentence in a vector, are maybe even more widely used in industry than generative models.
The workhorse for this category has been BERT since its release in 2018 (that's prehistory for LLMs).
It's not a fancy 100B parameters supermodel (just a few hundred millions), but it's an excellent workhorse, kind of a Honda Civic for LLMs.
Many applications use BERT-family models - the top models in this category cumulate millions of downloads on the Hub.
โก๏ธ Now a collaboration between Answer.AI and LightOn just introduced BERT's replacement: ModernBERT.
๐ง๐;๐๐ฅ:
๐๏ธ Architecture changes:
โ First, standard modernizations:
- Rotary positional embeddings (RoPE)
- Replace GeLU with GeGLU,
- Use Flash Attention 2
โจ The team also introduced innovative techniques like alternating attention instead of full attention, and sequence packing to get rid of padding overhead.
๐ฅ As a result, the model tops the game of encoder models:
It beats previous standard DeBERTaV3 for 1/5th the memory footprint, and runs 4x faster!
Read the blog post ๐ https://huggingface.co/blog/modernbert
We talk a lot about โจGenerative AIโจ, meaning "Decoder version of the Transformers architecture", but this is only one of the ways to build LLMs: encoder models, that turn a sentence in a vector, are maybe even more widely used in industry than generative models.
The workhorse for this category has been BERT since its release in 2018 (that's prehistory for LLMs).
It's not a fancy 100B parameters supermodel (just a few hundred millions), but it's an excellent workhorse, kind of a Honda Civic for LLMs.
Many applications use BERT-family models - the top models in this category cumulate millions of downloads on the Hub.
โก๏ธ Now a collaboration between Answer.AI and LightOn just introduced BERT's replacement: ModernBERT.
๐ง๐;๐๐ฅ:
๐๏ธ Architecture changes:
โ First, standard modernizations:
- Rotary positional embeddings (RoPE)
- Replace GeLU with GeGLU,
- Use Flash Attention 2
โจ The team also introduced innovative techniques like alternating attention instead of full attention, and sequence packing to get rid of padding overhead.
๐ฅ As a result, the model tops the game of encoder models:
It beats previous standard DeBERTaV3 for 1/5th the memory footprint, and runs 4x faster!
Read the blog post ๐ https://huggingface.co/blog/modernbert
- 1 reply
updated
a
Space
16 days ago
posted
an
update
16 days ago
Post
2282
๐๐ฎ๐ ๐ ๐ข๐ง๐ ๐
๐๐๐ ๐ซ๐๐ฅ๐๐๐ฌ๐๐ฌ ๐๐ข๐๐จ๐ญ๐ซ๐จ๐ง, ๐ ๐ฆ๐ข๐๐ซ๐จ๐ฌ๐๐จ๐ฉ๐ข๐ ๐ฅ๐ข๐ ๐ญ๐ก๐๐ญ ๐ฌ๐จ๐ฅ๐ฏ๐๐ฌ ๐๐๐ ๐ญ๐ซ๐๐ข๐ง๐ข๐ง๐ ๐๐ ๐ฉ๐๐ซ๐๐ฅ๐ฅ๐๐ฅ๐ข๐ณ๐๐ญ๐ข๐จ๐ง ๐ฅณ
๐ฐ๏ธ Llama-3.1-405B took 39 million GPU-hours to train, i.e. about 4.5 thousand years.
๐ด๐ป If they had needed all this time, we would have GPU stories from the time of Pharaoh ๐: "Alas, Lord of Two Lands, the shipment of counting-stones arriving from Cathay was lost to pirates, this shall delay the building of your computing temple by many moons "
๐ ๏ธ But instead, they just parallelized the training on 24k H100s, which made it take just a few months.
This required parallelizing across 4 dimensions: data, tensor, context, pipeline.
And it is infamously hard to do, making for bloated code repos that hold together only by magic.
๐ค ๐๐๐ ๐ป๐ผ๐ ๐๐ฒ ๐ฑ๐ผ๐ป'๐ ๐ป๐ฒ๐ฒ๐ฑ ๐ต๐๐ด๐ฒ ๐ฟ๐ฒ๐ฝ๐ผ๐ ๐ฎ๐ป๐๐บ๐ผ๐ฟ๐ฒ! Instead of building mega-training codes, Hugging Face colleagues cooked in the other direction, towards tiny 4D parallelism libs. A team has built Nanotron, already widely used in industry.
And now a team releases Picotron, a radical approach to code 4D Parallelism in just a few hundred lines of code, a real engineering prowess, making it much easier to understand what's actually happening!
โก ๐๐'๐ ๐๐ถ๐ป๐, ๐๐ฒ๐ ๐ฝ๐ผ๐๐ฒ๐ฟ๐ณ๐๐น:
Counting in MFU (Model FLOPs Utilization, how much the model actually uses all the compute potential), this lib reaches ~50% on SmolLM-1.7B model with 8 H100 GPUs, which is really close to what huge libs would reach. (Caution: the team is leading further benchmarks to verify this)
Go take a look ๐ https://github.com/huggingface/picotron/tree/main/picotron
๐ฐ๏ธ Llama-3.1-405B took 39 million GPU-hours to train, i.e. about 4.5 thousand years.
๐ด๐ป If they had needed all this time, we would have GPU stories from the time of Pharaoh ๐: "Alas, Lord of Two Lands, the shipment of counting-stones arriving from Cathay was lost to pirates, this shall delay the building of your computing temple by many moons "
๐ ๏ธ But instead, they just parallelized the training on 24k H100s, which made it take just a few months.
This required parallelizing across 4 dimensions: data, tensor, context, pipeline.
And it is infamously hard to do, making for bloated code repos that hold together only by magic.
๐ค ๐๐๐ ๐ป๐ผ๐ ๐๐ฒ ๐ฑ๐ผ๐ป'๐ ๐ป๐ฒ๐ฒ๐ฑ ๐ต๐๐ด๐ฒ ๐ฟ๐ฒ๐ฝ๐ผ๐ ๐ฎ๐ป๐๐บ๐ผ๐ฟ๐ฒ! Instead of building mega-training codes, Hugging Face colleagues cooked in the other direction, towards tiny 4D parallelism libs. A team has built Nanotron, already widely used in industry.
And now a team releases Picotron, a radical approach to code 4D Parallelism in just a few hundred lines of code, a real engineering prowess, making it much easier to understand what's actually happening!
โก ๐๐'๐ ๐๐ถ๐ป๐, ๐๐ฒ๐ ๐ฝ๐ผ๐๐ฒ๐ฟ๐ณ๐๐น:
Counting in MFU (Model FLOPs Utilization, how much the model actually uses all the compute potential), this lib reaches ~50% on SmolLM-1.7B model with 8 H100 GPUs, which is really close to what huge libs would reach. (Caution: the team is leading further benchmarks to verify this)
Go take a look ๐ https://github.com/huggingface/picotron/tree/main/picotron
- 1 reply
updated
a
Space
19 days ago
upvoted
an
article
19 days ago
Article
๐ช๐บโ๏ธ EU AI Act: Systemic Risks in the First CoP Draft Comments โ๏ธ๐ช๐บ
By
โข
โข
12