Models

Llama-3.2-3B-Instruct_new_alpaca_003

qwen3_1.7b_rush_hour_one_move_4_9_epoch1

qwen3_1.7b_rush_hour_one_move_4_9_epoch2

qwen3_1.7b_rush_hour_one_move_4_9

ds1p5b_code_sandbox-global_step_700

Qwen3-1.7B-Base_csum_6_10_geq_8_geq_8_0p75_0p5_1p0_0p0_1p0_grpo_42_rule

pdalma_ctx4_dm1_ce003_pr05_ptll32-1b_s2_ckpt_5_of_10_it36

ds1p5b_code_sandbox-global_step_800

pdalma_ctx4_dm1_ce0_pr1_ptll32-1b_s2_ckpt_1_of_10_it4

qwen3_1.7b_rush_hour_multi_move_final_10_12

ds1p5b_skywork_math_hard-global_step_200

ds1p5b_skywork_math_hard-global_step_400

qwen3_1.7b_sudoku_multi_action_easy_11_20_epoch3

qwen25-3b-l3l3-ep5

grpo_rmsprop_qwen3_1p7b_3k_seqlen_1e-6

grpo_rmsprop_qwen3_1p7b_3k_seqlen_1e-5

DAPO_GRPO_8b_incorrect_bs_32_mb_8_n16_cliphigh

gra4

Qwen2.5-0.5B-Instruct-Gensyn-Swarm-sturdy_finicky_cat

erpo-iclr-ours-Qwen2.5-3b-corr_gen_s002_max12

Qwen2.5-0.5B-Instruct-Gensyn-Swarm-peckish_stinging_macaque

giguan

CodeLlama3.2-3B-1225

qwen2.5-1.5b-pro

nvidia_math_cot_1e5_v2_ep5

sft_qwen15_code200_lr_1e-5_cosine_bsz_128_ckpt_2_of_5

sft_qwen15_code200_lr_5e-6_constant_bsz_128_ckpt_4_of_5

pdcd200_cptq15_ce003_pr05_ptq25-15b_omi_c100k_200tok_s8_ckpt_2_of_10_it26

pdalma_ctx4_dm1_ce01_pr0_ptll32-1b_s2_ckpt_2_of_10_it7

qwen3_1.7b_rush_hour_multi_move_final_short_4_9_epoch3

qwen-arc-abs-gpt5.2-sft-fewshot4-1epoch-icmlpaper-0125

SDRL-rand-Qwen3-4B-Base-icml-self-debate-random_n8_l2048-DAPO_n8_bs256_long8-step200

1_to_16_analysis

Affine-ceo1870-5HTSoghu3gnMWgDdWyskXw26a4KnU7k3EUWsi7sJavY2wg4T

DeepSeek-R1-Distill-Qwen-1.5B

qwen3_4b_sudoku_one_act_sft_final

Qwen3-1.7B-CCC-merged-cp4-LR1e-4

MNLP_M2_mcqa_model

old-122

fff-ooo

qwen3-8B-Base-orca_math-sparse-LoRA-step180-merged

paper_qwen_qwen3-instruct-4b_train_sft_all_train_think