DONGRYEOLLEE
drlee1
AI & ML interests
LLM Agents, RAG, Post-training, Efficient Inference & Reasoning, Mechanistic Interpretability
Recent Activity
updated a dataset about 15 hours ago
drlee1/hubsift-snapshots liked a model 1 day ago
LiquidAI/LFM2.5-24B-A2B upvoted an article 1 day ago
UltraBERT: One Encoder for Classification and RetrievalOrganizations
None yet
Question on the architecture ablation and 1M-context evaluation
2
#3 opened 3 months ago
by drlee1
The sample code for usage with Transformers is incorrect.
➕ 1
2
#45 opened over 2 years ago
by 
endNone
Setting `pad_token_id` to `eos_token_id`:128001 for open-end generation.
5
#23 opened over 2 years ago
by 
hmanju
Setting `pad_token_id` to `eos_token_id`:128001 for open-end generation.
5
#23 opened over 2 years ago
by hmanju