swerl_qwen3_8b_our_sft_tmax_10k_grpo_step500
668midterm-8bitFT
gptlong_continue_nemotron_terminal_step3300__Qwen3-32B
Qwen3-0.6B-Reverse-Text-SFT
Llama-3.2-3B-Instruct-hhrlhf
menochat-gemma3_4b-merged
bb1fe69d
safety_model
Affine-5ECFPTFqojMnEB6z881mJzrXLREvkEnj1wcu37zz4223Ln9x
multilingual_model
assn2-sft-llama32-1b
unified-model-stage1-5
qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step450
qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step50
qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step350
qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step580
qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step550
qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step300
qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step500
qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step450
qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step580
qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step400
qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step200
qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step250
qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step500
qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step400
qwen2.5_math_1.5b_grpo_prob_adv_scaled_ratio_w_o_kl_step350
ee_gol_grp_f1_form_spanOver
qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step550
ee_gol_grp_f1_form_wo_ns
qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step150
math_model
group_model
qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step250
f5bd0cc4
qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step200
qwen2.5_math_1.5b_grpo_rollout_8_w_o_KL_step300
Llama-3.1-8B-Instruct_grpo_ppl_adv_rollout_8_kl_0.001_20260516_140637_step232
grpo_adv_rollout_8_20260513_123609_USE_KL_True_step580
PureRL-1.5B-v6c1-distill-lam01-maskoff
sac-gspo-cl3e3-drgrpo-qwen25-math-1.5b-step1381
