DAPO: An Open-Source LLM Reinforcement Learning System at Scale
Paper
•
2503.14476
•
Published
•
104
Qiying Yu
qiying
AI & ML interests
None yet
Recent Activity
authored
a paper
6 days ago
DAPO: An Open-Source LLM Reinforcement Learning System at Scale
upvoted
a
paper
7 days ago
DAPO: An Open-Source LLM Reinforcement Learning System at Scale
updated
a dataset
8 days ago
BytedTsinghua-SIA/DAPO-Math-17k
Organizations
qiying's activity
authored
a
paper
6 days ago
upvoted
a
paper
7 days ago
updated
a
dataset
8 days ago
authored
a
paper
6 months ago
liked
a
model
7 months ago
About the Data Generation Method
2
#4 opened 7 months ago
by
qiying
liked
a
dataset
7 months ago
updated
2
models
8 months ago
liked
a
dataset
8 months ago
upvoted
a
collection
9 months ago
upvoted
an
article
9 months ago
Article
Cosmopedia: how to create large-scale synthetic data for pre-training Large Language Models
•
82
liked
a
Space
9 months ago
How do you use the rationales and answers in your training?
2
#1 opened 10 months ago
by
qiying
reacted to
merve's
post with ❤️
10 months ago
Post
EVA-CLIP 🦖 is the CLIP scaled to the moon! 🔥
The new SotA CLIP-like model 🏆
Highlights ✨
- Performs better in linear probing
- Outperforms in Zero-Shot Image-Text Retrieval
- Higher zero-shot accuracy in IN-1K
As usual, try it with the notebook I built for you https://colab.research.google.com/drive/1K7DdCORC3x4qyhwhuB4fT4wcfJ_BQLKw?usp=sharing#scrollTo=0ZS_lJ7SK6Ys
I also built a Space for you to compare the output probabilities to CLIP, seems that EVACLIP is more "sure" of it's results 😊 merve/EVACLIP
The authors have shared 8B checkpoints open with Apache 2.0 license 💜 and it's built on top of transformers, super easy to use! BAAI/EVA-CLIP-8B
Read the paper EVA-CLIP-18B: Scaling CLIP to 18 Billion Parameters (2402.04252) 📄
The new SotA CLIP-like model 🏆
Highlights ✨
- Performs better in linear probing
- Outperforms in Zero-Shot Image-Text Retrieval
- Higher zero-shot accuracy in IN-1K
As usual, try it with the notebook I built for you https://colab.research.google.com/drive/1K7DdCORC3x4qyhwhuB4fT4wcfJ_BQLKw?usp=sharing#scrollTo=0ZS_lJ7SK6Ys
I also built a Space for you to compare the output probabilities to CLIP, seems that EVACLIP is more "sure" of it's results 😊 merve/EVACLIP
The authors have shared 8B checkpoints open with Apache 2.0 license 💜 and it's built on top of transformers, super easy to use! BAAI/EVA-CLIP-8B
Read the paper EVA-CLIP-18B: Scaling CLIP to 18 Billion Parameters (2402.04252) 📄
How do you use the rationales and answers in the arxivqa training?
#1 opened 10 months ago
by
qiying
liked
a
model
11 months ago
liked
a
model
12 months ago
liked
a
Space
about 1 year ago
Training Hyperparameters
#6 opened about 1 year ago
by
qiying