arxiv:2506.06632
Shubham Parashar
shubhamprshr
AI & ML interests
Computer Vision, Multi-Modal Learning
Recent Activity
authored a paper 5 days ago
Curriculum Reinforcement Learning from Easy to Hard Tasks Improves LLM Reasoning updated a model 6 days ago
divelab/DAPO_E2H-countdown-gaussian_0p5_0p5 updated a model 6 days ago
divelab/DAPO_E2H-math-gaussian_0p5_0p5Organizations
models 93
shubhamprshr/Qwen2.5-1.5B-Instruct_gsm8k_grpo_gaussian_0.5_0.5_SEC0.3DRO1.0G0.0_minpTrue_1600
Text Generation • 2B • Updated • 4
shubhamprshr/Qwen2.5-1.5B-Instruct_gsm8k_grpo_gaussian_0.25_0.75_SEC0.3DRO1.0G0.0_minpTrue_1600
Text Generation • 2B • Updated • 2
shubhamprshr/Qwen2.5-1.5B-Instruct_countdown2345_grpo_cosine_0.5_0.5_SEC0.3DRO1.0G0.0_minpTrue_1600
Text Generation • 2B • Updated • 3
shubhamprshr/Qwen2.5-1.5B-Instruct_countdown2345_grpo_gaussian_0.5_0.5_SEC0.3DRO1.0G0.0_minpTrue_1600
Text Generation • 2B • Updated • 2
shubhamprshr/Qwen2.5-1.5B-Instruct_math_grpo_gaussian_0.5_0.5_SEC0.3DRO1.0G0.0_minpTrue_1600
Text Generation • 2B • Updated • 2
shubhamprshr/Qwen2.5-1.5B-Instruct_math_grpo_cosine_0.5_0.5_SEC0.3DRO1.0G0.0_minpTrue_1600
Text Generation • 2B • Updated • 2
shubhamprshr/Qwen2.5-1.5B-Instruct_math_grpo_balanced_0.5_0.5_SEC0.3DRO1.0G0.0_minpTrue_1600
Text Generation • 2B • Updated • 2
shubhamprshr/Qwen2.5-1.5B-Instruct_countdown2345_grpo_balanced_0.5_0.5_SEC0.3DRO1.0G0.0_minpTrue_1600
Text Generation • 2B • Updated • 2
shubhamprshr/Qwen2.5-1.5B-Instruct_blocksworld1246_grpo_cosine_0.5_0.5_SEC0.3DRO1.0G0.0_minpTrue_1200
Text Generation • 2B • Updated • 4
shubhamprshr/Qwen2.5-1.5B-Instruct_blocksworld1246_grpo_gaussian_0.25_0.75_SEC0.3DRO1.0G0.0_minpTrue_1200
Text Generation • 2B • Updated • 2
datasets 0
None public yet