Article
Deep Learning GPU Benchmarks
β’
2
Tudor M
tudorizer
AI & ML interests
Hardware, GPUs, renewable energy; enverge.ai
Recent Activity
upvoted
an
article
about 2 months ago
Deep Learning GPU Benchmarks
updated
a model
6 months ago
tudorizer/distilroberta-ai-job-embeddings
published
a model
6 months ago
tudorizer/distilroberta-ai-job-embeddings
Organizations
upvoted
an
article
about 2 months ago
updated
a
model
6 months ago
published
a
model
6 months ago
updated
a
dataset
6 months ago
published
a
dataset
6 months ago
upvoted
an
article
6 months ago
Article
Deploy LLMs with Hugging Face Inference Endpoints
β’
17
published
a
Space
8 months ago
upvoted
an
article
12 months ago
Article
Energy Scores for AI Models
β’
38
reacted to
m-ric's
post with π₯
about 1 year ago
Post
2608
After 6 years, BERT, the workhorse of encoder models, finally gets a replacement: πͺπ²πΉπ°πΌπΊπ² π πΌπ±π²πΏπ»πππ₯π§! π€
We talk a lot about β¨Generative AIβ¨, 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.
β‘οΈ Now a collaboration between Answer.AI and LightOn just introduced BERT's replacement: ModernBERT.
π§π;ππ₯:
ποΈ Architecture changes:
β First, standard modernizations:
- Rotary positional embeddings (RoPE)
- Replace GeLU with GeGLU,
- Use Flash Attention 2
β¨ The team also introduced innovative techniques like alternating attention instead of full attention, and sequence packing to get rid of padding overhead.
π₯ 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 π https://huggingface.co/blog/modernbert
We talk a lot about β¨Generative AIβ¨, 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.
β‘οΈ Now a collaboration between Answer.AI and LightOn just introduced BERT's replacement: ModernBERT.
π§π;ππ₯:
ποΈ Architecture changes:
β First, standard modernizations:
- Rotary positional embeddings (RoPE)
- Replace GeLU with GeGLU,
- Use Flash Attention 2
β¨ The team also introduced innovative techniques like alternating attention instead of full attention, and sequence packing to get rid of padding overhead.
π₯ 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 π https://huggingface.co/blog/modernbert
- 1 reply
updated
a
Space
about 1 year ago
updated
3
models
about 1 year ago