Yucheng
liyucheng
AI & ML interests
Robust LLMs Evaluation and Efficient LLMs Inference.
Recent Activity
liked
a dataset
5 days ago
YangZhoumill/factorreit_32k_hard
liked
a dataset
5 days ago
YangZhoumill/factorreit_32k_medium
liked
a dataset
5 days ago
YangZhoumill/factorreit_16k_hard
Organizations
liyucheng's activity
Converting to native Transformers
13
#17 opened 4 months ago
by
cyrilvallez
![](https://cdn-avatars.huggingface.co/v1/production/uploads/649a943906690b959319d192/k8ZlmJWF3GNpYLHrdpejY.jpeg)
tokenizer hf 4.46.3
#19 opened 2 months ago
by
liyucheng
![](https://cdn-avatars.huggingface.co/v1/production/uploads/63d00710645dd8d34ea9bcc6/E6YsIsXH57OACL-NZ52fB.jpeg)
multi-gpu encode?
#47 opened 2 months ago
by
liyucheng
![](https://cdn-avatars.huggingface.co/v1/production/uploads/63d00710645dd8d34ea9bcc6/E6YsIsXH57OACL-NZ52fB.jpeg)
Relation with Marco-o1
1
#10 opened 2 months ago
by
Viper403
error with 2024-09
2
#4 opened 2 months ago
by
sewon
![](https://cdn-avatars.huggingface.co/v1/production/uploads/63a76d0de27a6dbd485fe863/qJJwHOuvyQGq1o0KscOF_.jpeg)
Limitation - context length of the BERT classifier?
2
#17 opened 2 months ago
by
liyucheng
![](https://cdn-avatars.huggingface.co/v1/production/uploads/63d00710645dd8d34ea9bcc6/E6YsIsXH57OACL-NZ52fB.jpeg)
Choice on pretrained model and fine-tuning.
2
#3 opened 8 months ago
by
Avditvs
classification or regression? how to choose.
1
#7 opened 3 months ago
by
DaleMeng
Hi, very cool dataset. Can you please grant me the access of this dataset? Many thanks!
#2 opened 3 months ago
by
liyucheng
![](https://cdn-avatars.huggingface.co/v1/production/uploads/63d00710645dd8d34ea9bcc6/E6YsIsXH57OACL-NZ52fB.jpeg)
Bug: Model fails on Long-context tests that the previous version can pass.
3
#1 opened 3 months ago
by
liyucheng
![](https://cdn-avatars.huggingface.co/v1/production/uploads/63d00710645dd8d34ea9bcc6/E6YsIsXH57OACL-NZ52fB.jpeg)
Hi Markus, nice work. Is it possible to grant me the access of this dataset? Many thanks.
#3 opened 3 months ago
by
liyucheng
![](https://cdn-avatars.huggingface.co/v1/production/uploads/63d00710645dd8d34ea9bcc6/E6YsIsXH57OACL-NZ52fB.jpeg)
Cool dataset! Do we have a pure textual version jailbreaks provided?
1
#3 opened 3 months ago
by
liyucheng
![](https://cdn-avatars.huggingface.co/v1/production/uploads/63d00710645dd8d34ea9bcc6/E6YsIsXH57OACL-NZ52fB.jpeg)
The first 1500 chars are used instead of 1000.
#4 opened 3 months ago
by
liyucheng
![](https://cdn-avatars.huggingface.co/v1/production/uploads/63d00710645dd8d34ea9bcc6/E6YsIsXH57OACL-NZ52fB.jpeg)
Hi Luca, what's the source of this dataset?
3
#1 opened 3 months ago
by
liyucheng
![](https://cdn-avatars.huggingface.co/v1/production/uploads/63d00710645dd8d34ea9bcc6/E6YsIsXH57OACL-NZ52fB.jpeg)
Data from 2024-06 to 2024-09 has very few rows
2
#3 opened 4 months ago
by
jijivski
Converting to native Transformers
22
#81 opened 4 months ago
by
cyrilvallez
![](https://cdn-avatars.huggingface.co/v1/production/uploads/649a943906690b959319d192/k8ZlmJWF3GNpYLHrdpejY.jpeg)