Benchmarking LLM capability of external and internal error correction
Ken Tsui
kenhktsui
·
AI & ML interests
ML engineer, researcher
VLM, LLM benchmark
Opinions are my own
Recent Activity
upvoted a paper about 1 month ago
A Very Big Video Reasoning Suite liked a model 2 months ago
moonshotai/Kimi-K2.5 liked a dataset 3 months ago
VITRA-VLA/VITRA-1MOrganizations
LongTalk
A Very Long Chain-of-Thought Dataset for Reasoning Model Post-Training
-
kenhktsui/longtalk-cot-v0.1
Viewer • Updated • 61.2k • 104 • 13 -
kenhktsui/qwen2.5-7b-instruct-thinking-sft-merged-gguf
8B • Updated • 17 • 1 -
kenhktsui/qwen2.5-7b-instruct-thinking-sft-merged
Text Generation • 8B • Updated • 2 -
kenhktsui/llama3.1-8b-instruct-thinking-sft-merged-gguf
8B • Updated • 52 • 1
CoT
VLM Data
FastText Model for Pretraining Data Curation
-
kenhktsui/llm-data-textbook-quality-fasttext-classifier-v2
Text Classification • Updated • 215 • 28 -
kenhktsui/fineweb-edu-fasttext-classifier
Text Classification • Updated • 4 • 4 -
kenhktsui/code-natural-language-fasttext-classifier
Text Classification • Updated • 407 • 5 -
kenhktsui/math-fasttext-classifier
Text Classification • Updated • 14 • 2
textbook-quality-classifier
-
kenhktsui/fineweb-edu-fasttext-classifier
Text Classification • Updated • 4 • 4 -
kenhktsui/llm-data-textbook-quality-fasttext-classifier-v2
Text Classification • Updated • 215 • 28 -
kenhktsui/llm-data-textbook-quality-classifier-v1
Text Classification • 0.3B • Updated • 15 • 10 -
kenhktsui/llm-data-textbook-quality-fasttext-classifier-v1
Text Classification • Updated • 13 • 4
nano-phi
Small Language Model Trained with Textbook Quality Data - How Far Can It Go?
Self Correction Bench
Benchmarking LLM capability of external and internal error correction
FastText Model for Pretraining Data Curation
-
kenhktsui/llm-data-textbook-quality-fasttext-classifier-v2
Text Classification • Updated • 215 • 28 -
kenhktsui/fineweb-edu-fasttext-classifier
Text Classification • Updated • 4 • 4 -
kenhktsui/code-natural-language-fasttext-classifier
Text Classification • Updated • 407 • 5 -
kenhktsui/math-fasttext-classifier
Text Classification • Updated • 14 • 2
LongTalk
A Very Long Chain-of-Thought Dataset for Reasoning Model Post-Training
-
kenhktsui/longtalk-cot-v0.1
Viewer • Updated • 61.2k • 104 • 13 -
kenhktsui/qwen2.5-7b-instruct-thinking-sft-merged-gguf
8B • Updated • 17 • 1 -
kenhktsui/qwen2.5-7b-instruct-thinking-sft-merged
Text Generation • 8B • Updated • 2 -
kenhktsui/llama3.1-8b-instruct-thinking-sft-merged-gguf
8B • Updated • 52 • 1
textbook-quality-classifier
-
kenhktsui/fineweb-edu-fasttext-classifier
Text Classification • Updated • 4 • 4 -
kenhktsui/llm-data-textbook-quality-fasttext-classifier-v2
Text Classification • Updated • 215 • 28 -
kenhktsui/llm-data-textbook-quality-classifier-v1
Text Classification • 0.3B • Updated • 15 • 10 -
kenhktsui/llm-data-textbook-quality-fasttext-classifier-v1
Text Classification • Updated • 13 • 4
CoT
nano-phi
Small Language Model Trained with Textbook Quality Data - How Far Can It Go?
VLM Data