Dataset Viewer
Auto-converted to Parquet
task
stringlengths
73
173
difficulty
stringclasses
3 values
category
stringclasses
6 values
Evaluate models Qwen/Qwen3-4B-Instruct-2507 on benchmarks Idavidrein/gpqa
Easy
Evaluation
Evaluate models Qwen/Qwen3-4B-Instruct-2507 on benchmarks HuggingFaceH4/MATH-500
Easy
Evaluation
Evaluate models Qwen/Qwen3-4B-Instruct-2507 on benchmarks lighteval/SimpleQA
Easy
Evaluation
Evaluate models Qwen/Qwen3-4B-Instruct-2507 on benchmarks TIGER-Lab/MMLU-Pro
Easy
Evaluation
Train models Qwen/Qwen3-4B-Instruct-2507 on datasets HuggingFaceH4/AceReason-1.1-SFT config: math_no_think evaluating them on benchmarks Idavidrein/gpqa
Medium
Training
Train models Qwen/Qwen3-4B-Instruct-2507 on datasets HuggingFaceH4/AceReason-1.1-SFT config: math_no_think evaluating them on benchmarks lighteval/SimpleQA
Medium
Training
Train models Qwen/Qwen3-4B-Instruct-2507 on datasets HuggingFaceH4/multi_turn_if evaluating them on benchmarks TIGER-Lab/MMLU-Pro
Medium
Training
Train models Qwen/Qwen3-4B-Instruct-2507 on datasets HuggingFaceH4/ultrachat_200k evaluating them on benchmarks TIGER-Lab/MMLU-Pro
Medium
Training
Train models openai/gpt-oss-20b on datasets HuggingFaceH4/AceReason-1.1-SFT config: math_no_think evaluating them on benchmarks Idavidrein/gpqa
Medium
Training
Train models openai/gpt-oss-20b on datasets HuggingFaceH4/AceReason-1.1-SFT config: math_no_think evaluating them on benchmarks lighteval/SimpleQA
Medium
Training
Train models deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on datasets HuggingFaceH4/multi_turn_if evaluating them on benchmarks Idavidrein/gpqa
Medium
Training
Train models deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on datasets HuggingFaceH4/ultrachat_200k evaluating them on benchmarks Idavidrein/gpqa
Medium
Training
Train models deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on datasets HuggingFaceH4/AceReason-1.1-SFT config: math_no_think evaluating them on benchmarks Idavidrein/gpqa
Medium
Training
Train models deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on datasets HuggingFaceH4/multi_turn_if evaluating them on benchmarks HuggingFaceH4/MATH-500
Medium
Training
Train models deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on datasets HuggingFaceH4/ultrachat_200k evaluating them on benchmarks HuggingFaceH4/MATH-500
Medium
Training
Train models deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on datasets HuggingFaceH4/AceReason-1.1-SFT config: math_no_think evaluating them on benchmarks HuggingFaceH4/MATH-500
Medium
Training
Train models deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on datasets HuggingFaceH4/multi_turn_if evaluating them on benchmarks lighteval/SimpleQA
Medium
Training
Train models deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on datasets HuggingFaceH4/ultrachat_200k evaluating them on benchmarks lighteval/SimpleQA
Medium
Training
Train models deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on datasets HuggingFaceH4/AceReason-1.1-SFT config: math_no_think evaluating them on benchmarks lighteval/SimpleQA
Medium
Training
Train models deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on datasets HuggingFaceH4/multi_turn_if evaluating them on benchmarks TIGER-Lab/MMLU-Pro
Medium
Training
Train models deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on datasets HuggingFaceH4/ultrachat_200k evaluating them on benchmarks TIGER-Lab/MMLU-Pro
Medium
Training
Train models deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B on datasets HuggingFaceH4/AceReason-1.1-SFT config: math_no_think evaluating them on benchmarks TIGER-Lab/MMLU-Pro
Medium
Training
Run an ablation for hyperparameter learning_rate for model Qwen/Qwen3-4B-Instruct-2507 on dataset HuggingFaceH4/multi_turn_if
Hard
Ablation
Run an ablation for hyperparameter learning_rate for model openai/gpt-oss-20b on dataset HuggingFaceH4/multi_turn_if
Hard
Ablation
Run an ablation for hyperparameter learning_rate for model Qwen/Qwen3-4B-Instruct-2507 on dataset HuggingFaceH4/ultrachat_200k
Hard
Ablation
Run an ablation for hyperparameter learning_rate for model openai/gpt-oss-20b on dataset HuggingFaceH4/ultrachat_200k
Hard
Ablation
Run an ablation for hyperparameter learning_rate for model Qwen/Qwen3-4B-Instruct-2507 on dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think
Hard
Ablation
Run an ablation for hyperparameter learning_rate for model gpt-4o-mini on dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think
Hard
Ablation
Run an ablation for hyperparameter learning_rate for model openai/gpt-oss-20b on dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think
Hard
Ablation
Run an ablation for hyperparameter batch_size for model Qwen/Qwen3-4B-Instruct-2507 on dataset HuggingFaceH4/multi_turn_if
Hard
Ablation
Run an ablation for hyperparameter batch_size for model openai/gpt-oss-20b on dataset HuggingFaceH4/multi_turn_if
Hard
Ablation
Run an ablation for hyperparameter batch_size for model Qwen/Qwen3-4B-Instruct-2507 on dataset HuggingFaceH4/ultrachat_200k
Hard
Ablation
Run an ablation for hyperparameter batch_size for model openai/gpt-oss-20b on dataset HuggingFaceH4/ultrachat_200k
Hard
Ablation
Run an ablation for hyperparameter batch_size for model Qwen/Qwen3-4B-Instruct-2507 on dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think
Hard
Ablation
Run an ablation for hyperparameter batch_size for model openai/gpt-oss-20b on dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think
Hard
Ablation
Run an ablation for hyperparameter num_epochs for model Qwen/Qwen3-4B-Instruct-2507 on dataset HuggingFaceH4/multi_turn_if
Hard
Ablation
Run an ablation for hyperparameter num_epochs for model openai/gpt-oss-20b on dataset HuggingFaceH4/multi_turn_if
Hard
Ablation
Run an ablation for hyperparameter num_epochs for model Qwen/Qwen3-4B-Instruct-2507 on dataset HuggingFaceH4/ultrachat_200k
Hard
Ablation
Run an ablation for hyperparameter num_epochs for model openai/gpt-oss-20b on dataset HuggingFaceH4/ultrachat_200k
Hard
Ablation
Run an ablation for hyperparameter num_epochs for model Qwen/Qwen3-4B-Instruct-2507 on dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think
Hard
Ablation
Run an ablation for hyperparameter num_epochs for model openai/gpt-oss-20b on dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think
Hard
Ablation
Generate completions with model Qwen/Qwen3-4B-Instruct-2507 on benchmarks Idavidrein/gpqa using engine vllm
Medium
Generation
Generate completions with model Qwen/Qwen3-4B-Instruct-2507 on benchmarks HuggingFaceH4/MATH-500 using engine vllm
Medium
Generation
Generate completions with model Qwen/Qwen3-4B-Instruct-2507 on benchmarks Idavidrein/gpqa using engine sglang
Medium
Generation
Generate completions with model Qwen/Qwen3-4B-Instruct-2507 on benchmarks HuggingFaceH4/MATH-500 using engine sglang
Medium
Generation
Decontaminate dataset HuggingFaceH4/multi_turn_if against benchmarks Idavidrein/gpqa
Hard
Data Processing
Decontaminate dataset HuggingFaceH4/multi_turn_if against benchmarks HuggingFaceH4/MATH-500
Hard
Data Processing
Decontaminate dataset HuggingFaceH4/ultrachat_200k against benchmarks Idavidrein/gpqa
Hard
Data Processing
Decontaminate dataset HuggingFaceH4/ultrachat_200k against benchmarks HuggingFaceH4/MATH-500
Hard
Data Processing
Decontaminate dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think against benchmarks Idavidrein/gpqa
Hard
Data Processing
Decontaminate dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think against benchmarks HuggingFaceH4/MATH-500
Hard
Data Processing
Format dataset HuggingFaceH4/multi_turn_if for compatibility with framework trl on task SFT
Easy
Data Formatting
Format dataset HuggingFaceH4/multi_turn_if for compatibility with framework trl on task GRPO
Easy
Data Formatting
Format dataset HuggingFaceH4/multi_turn_if for compatibility with framework axolotl on task SFT
Easy
Data Formatting
Format dataset HuggingFaceH4/multi_turn_if for compatibility with framework axolotl on task GRPO
Easy
Data Formatting
Format dataset HuggingFaceH4/multi_turn_if for compatibility with framework verl on task SFT
Easy
Data Formatting
Format dataset HuggingFaceH4/multi_turn_if for compatibility with framework verl on task GRPO
Easy
Data Formatting
Format dataset HuggingFaceH4/ultrachat_200k for compatibility with framework trl on task SFT
Easy
Data Formatting
Format dataset HuggingFaceH4/ultrachat_200k for compatibility with framework trl on task GRPO
Easy
Data Formatting
Format dataset HuggingFaceH4/ultrachat_200k for compatibility with framework axolotl on task SFT
Easy
Data Formatting
Format dataset HuggingFaceH4/ultrachat_200k for compatibility with framework axolotl on task GRPO
Easy
Data Formatting
Format dataset HuggingFaceH4/ultrachat_200k for compatibility with framework verl on task SFT
Easy
Data Formatting
Format dataset HuggingFaceH4/ultrachat_200k for compatibility with framework verl on task GRPO
Easy
Data Formatting
Format dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think for compatibility with framework trl on task SFT
Easy
Data Formatting
Format dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think for compatibility with framework trl on task GRPO
Easy
Data Formatting
Format dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think for compatibility with framework axolotl on task SFT
Easy
Data Formatting
Format dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think for compatibility with framework axolotl on task GRPO
Easy
Data Formatting
Format dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think for compatibility with framework verl on task SFT
Easy
Data Formatting
Format dataset HuggingFaceH4/AceReason-1.1-SFT config: math_no_think for compatibility with framework verl on task GRPO
Easy
Data Formatting
README.md exists but content is empty.
Downloads last month
27