On-Policy Distillation Analysis . yunjae-won/qwen1.7b_clip1e-6_base_step150 Text Generation • 2B • Updated 26 days ago • 44 yunjae-won/qwen1.7b_clip1e-6_base_step125 Text Generation • 2B • Updated 26 days ago • 33 yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_staticKL_reg0.5_step150 Text Generation • 4B • Updated 3 days ago • 28 yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_staticKL_reg0.5_step125 Text Generation • 4B • Updated 3 days ago • 33
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_staticKL_reg0.5_step150 Text Generation • 4B • Updated 3 days ago • 28
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_staticKL_reg0.5_step125 Text Generation • 4B • Updated 3 days ago • 33
dpo-info-loss yunjae-won/mpq3_qwen4bi_sft Text Generation • 4B • Updated Apr 7 • 6 yunjae-won/mpq3_qwen4bi_sft_dpo_beta1e-1_step256 Text Generation • 4B • Updated Apr 6 • 5 yunjae-won/mpq3_qwen4bi_sft_dpo_beta1e-1_step512 Text Generation • 4B • Updated Apr 6 • 11 yunjae-won/mpq3_qwen4bi_sft_dpo_beta1e-1_step768 Text Generation • 4B • Updated Apr 6 • 6
On-Policy Distillation Analysis . yunjae-won/qwen1.7b_clip1e-6_base_step150 Text Generation • 2B • Updated 26 days ago • 44 yunjae-won/qwen1.7b_clip1e-6_base_step125 Text Generation • 2B • Updated 26 days ago • 33 yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_staticKL_reg0.5_step150 Text Generation • 4B • Updated 3 days ago • 28 yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_staticKL_reg0.5_step125 Text Generation • 4B • Updated 3 days ago • 33
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_staticKL_reg0.5_step150 Text Generation • 4B • Updated 3 days ago • 28
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_staticKL_reg0.5_step125 Text Generation • 4B • Updated 3 days ago • 33
dpo-info-loss yunjae-won/mpq3_qwen4bi_sft Text Generation • 4B • Updated Apr 7 • 6 yunjae-won/mpq3_qwen4bi_sft_dpo_beta1e-1_step256 Text Generation • 4B • Updated Apr 6 • 5 yunjae-won/mpq3_qwen4bi_sft_dpo_beta1e-1_step512 Text Generation • 4B • Updated Apr 6 • 11 yunjae-won/mpq3_qwen4bi_sft_dpo_beta1e-1_step768 Text Generation • 4B • Updated Apr 6 • 6
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_step25 Text Generation • 4B • Updated 2 days ago • 23
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_step200 Text Generation • 4B • Updated 2 days ago • 23
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_step175 Text Generation • 4B • Updated 2 days ago • 16
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_step100 Text Generation • 4B • Updated 2 days ago • 22
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_step50 Text Generation • 4B • Updated 2 days ago • 17
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_step75 Text Generation • 4B • Updated 2 days ago • 24
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_step125 Text Generation • 4B • Updated 2 days ago • 24
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_step150 Text Generation • 4B • Updated 2 days ago • 23
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_adaKL_reg4_step25 Text Generation • 4B • Updated 2 days ago • 26
yunjae-won/qwen4b_noclip_default_KLeff_smooth_gamma1e-6_ReWeight_adaKL_reg4_step200 Text Generation • 4B • Updated 2 days ago • 25