2-3B Parameter LLMs — Page 54

agentic-sudoku-NoStateTrans_qwen2.5-3B-5e-6_gt-SFT_ans1-24k

qwen3_1.7b_sudoku_multi_action_easy_21_30_epoch3

north_llama32_3b_enhancedNCC_instruct_v1_long_large_lr2e6_2048_90000

TFRank-GRPO-Qwen3-1.7B

CORE-Qwen3-1.7B-MATH

qwen3_1.7b_easy_rl_reinforce_ori

qwen3_1.7b_sudoku_multi_action_easy_21_30_epoch2

qwen3_1.7b_sudoku_multi_action_easy_21_30_epoch1

qwen3_1.7b_sudoku_multi_action_easy_11_20

qwen3_1.7b_new_sudoku_one_action_B_sft_lr_5e_6__step_4432

Affine-at02-12-31-02

qwen3_1.7b_new_sudoku_one_action_C_sft_lr_5e_6__step_3336

final-d2-1.7b

qwen3_1.7b_rush_hour_multi_move_final_new

nvidia_math_cot_qwq_1e5

qwen3_1.7b_new_sudoku_one_action_C_sft_lr_5e_6__step_1668

qwen3_1.7b_new_sudoku_one_action_C_sft_lr_5e_6__step_6672

tool_cor_3B

qwen3_1.7b_sudoku_multi_action_easy_11_20_epoch1

qwen3_1.7b_sudoku_one_action_easy_11_20_epoch1

nvidia_qwq_aug_1e5

4b_SFT_NEW

qwen3-1.7b-base-svd-muon-adam-1e-6-bs128-kl0.0-global_step_200

evolved_set1_correct_12k_ep10

qwen3-1.7b-base-adam-2e-6-bs128-kl0.0-global_step_200

agentic-futoshiki-NonMarkov_qwen2.5-3B-5e-6_gt-SFT_20k

self-debate-baseline-Qwen3-1.7B-Base-DAPO-n8-bs256-long8-step200

agentic-futoshiki-NonMarkov_qwen2.5-3B-5e-6_gt-SFT_10k

SFT-Warmup-3B

agentic-sudoku-NonMarkov_qwen2.5-3B-5e-6_9x9_6-6_gt-SFT_ans1-7k

nvidia_math_cot_1e5_v2_ep10

qwen3-1.7b-base-svd-muon-adam-2e-6-bs128-kl0.0-global_step_200

qwen3-1.7b-base-svd-muon-adam-1e-6-bs128-kl0.0-global_step_100

CORE-Qwen3-1.7B-MATH-A9-U-S-SG

qwen3_1.7b_rush_hour_one_move_4_9_epoch1

qwen3_1.7b_sudoku_one_action_easy_21_30_epoch2

qwen3_1.7b_sudoku_one_action_easy_21_30_epoch3

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

qwen3_1.7b_rush_hour_multi_move_final_4_9_long_10_12_epoch3

dl_finetuned_minicoder

qwen2.5-3b-deep-research

qwen3-1.7b-amr-20260206-1038-1epoch