ldwang's picture

ldwang

ldwang

·

ftgreat

AI & ML interests

LLM, MLLM, Infra

Recent Activity

new activity about 21 hours ago

nvidia/ClimbMix:About MMLU metric in CLIMB paper

liked a model 1 day ago

Qwen/Qwen2.5-0.5B

liked a dataset 1 day ago

nvidia/ClimbMix

View all activity

Organizations

ldwang's activity

New activity in nvidia/ClimbMix about 21 hours ago

About MMLU metric in CLIMB paper

#4 opened 1 day ago by

commented a paper 18 days ago

Trajectory Balance with Asynchrony: Decoupling Exploration and Learning for Fast, Scalable LLM Post-Training

Paper • 2503.18929 • Published 27 days ago • 3 •

New activity in BAAI/Infinity-MM 25 days ago

数据集中有的图片是jpg，但实际上是png

#13 opened 25 days ago by

New activity in HuggingFaceTB/SmolLM2-1.7B-Instruct 5 months ago

Performance on MATH dataset?

#3 opened 6 months ago by

New activity in TRI-ML/DCLM-1B-v0 6 months ago

About datasets and training setting

#6 opened 6 months ago by

commented a paper 6 months ago

Infinity-MM: Scaling Multimodal Performance with Large-Scale and High-Quality Instruction Data

Paper • 2410.18558 • Published Oct 24, 2024 • 20 •

New activity in BAAI/CCI3-HQ 6 months ago

[bot] Conversion to Parquet

#1 opened 7 months ago by

parquet-converter

New activity in BAAI/CCI3-Data 6 months ago

What zh/en benchmarks are used for evaluation?

#1 opened 7 months ago by

New activity in BAAI/CCI3-HQ-Classifier 6 months ago

举的例子挺有意思的🙂

#1 opened 6 months ago by

New activity in ldwang/OpenHermes-2.5-zh 6 months ago

Dataset Viewer issue: DatasetGenerationCastError

#1 opened 8 months ago by

commented a paper 6 months ago

CCI3.0-HQ: a large-scale Chinese dataset of high quality designed for pre-training large language models

Paper • 2410.18505 • Published Oct 24, 2024 • 11 •

New activity in mamachang/medical-reasoning 6 months ago

About method

#2 opened 6 months ago by

New activity in HuggingFaceTB/SmolLM-1.7B 8 months ago

MMLU doesn't match on lm-evaluation-harness

#2 opened 9 months ago by

Adding Evaluation Results

#3 opened 9 months ago by

leaderboard-pr-bot

New activity in HuggingFaceFW/fineweb-edu-llama3-annotations 10 months ago

Should the data be filtered to score values between 1-5?

#1 opened 11 months ago by

New activity in BAAI/Aquila2-7B 11 months ago

Delete pytorch_model-00002-of-00002.bin

#1 opened 11 months ago by

New activity in facebook/galactica-30b about 2 years ago

metaseq checkpoint format

#10 opened about 2 years ago by