2-3B Parameter LLMs — Page 56
5,615MultiRLWarmTools2B32K
qwen3_1.7b_easy_rl_ours_adv_fixed_sequence_batch_128
MultiRLWarmTools2B32K
qwen3_1.7b_sft_final_easy_reinforce_ours_adv_fixed_gamma_0.9
MultiRLWarmTools2B32K
qwen3_1.7b_easy_rl_ours_adv_fixed_gamma_995_98_ori_norm
MultiRLWarmTools2B32K
qwen3_1.7b_easy_rl_ours_adv_fixed_geo_ms_only
MultiRLWarmTools2B32K
qwen3_1.7b_new_standard_A_sft_overfit_lr_5e_6__global_step_192
MultiRLWarmTools2B32K
qwen3_1.7b_new_standard_C_sft_overfit_lr_5e_5__global_step_1480
MultiRLWarmTools2B32K
qwen3_1.7b_new_standard_C_sft_overfit_lr_5e_5__global_step_1776
MultiRLWarmTools2B32K
qwen3_1.7b_new_standard_C_sft_overfit_lr_5e_5__global_step_888
MultiRLWarmTools2B32K
qwen3_1.7b_new_standard_C_sft_overfit_lr_5e_5__global_step_592
MultiRLWarmTools2B32K
qwen3_1.7b_new_standard_A_sft_overfit_lr_5e_6__global_step_288
MultiRLWarmTools2B32K
qwen3_1.7b_new_standard_A_sft_overfit_lr_5e_6__global_step_384
MultiRLWarmTools2B32K
qwen3_1.7b_new_standard_C_sft_overfit_lr_5e_6__global_step_1480
MultiRLWarmTools2B32K
qwen3_1.7b_new_standard_C_sft_overfit_lr_5e_6__global_step_1184
MultiRLWarmTools2B32K
qwen3_1.7b_new_standard_C_sft_overfit_lr_5e_6__global_step_296
MultiRLWarmTools2B32K
qwen3_1.7b_sudoku_one_action_easy_11_20
ericoh929WarmTools2B32K
qwen3-1.7b-huggingfaceh4-instruction-data-lora-instruction-tuned
MultiRLWarmTools2B32K
qwen3_1.7b_new_sudoku_one_action_A_sft_lr_5e_6__step_1124
MultiRLWarmTools2B32K
qwen3_1.7b_new_sudoku_one_action_A_sft_lr_5e_6__step_562
yurunyWarmTools3B32K
agentic-futoshiki-Markov_qwen2.5-3B-5e-6_gt-SFT_10k
MultiRLWarmTools2B32K
qwen3_1.7b_rush_hour_multi_move_sft_new
MultiRLWarmTools2B32K
qwen3_1.7b_one_act_easy_short
MultiRLWarmTools2B32K
qwen3_1.7b_rush_hour_one_move_4_9_epoch2
MultiRLWarmTools2B32K
qwen3_1.7b_rush_hour_one_move_4_9_epoch3
MultiRLWarmTools2B32K
qwen3_1.7b_rush_hour_one_move_4_9
MultiRLWarmTools2B32K
qwen3_1.7b_rush_hour_multi_move_final_10_12
MultiRLWarmTools2B32K
qwen3_1.7b_sudoku_multi_action_easy_11_20_epoch3
reds0510WarmTools3B32K
nvidia_math_cot_1e5_v2_ep5
JamesSandWarmTools2B32K
qwen1.7b-adam-reset-muon-lr-1e-6-fp64-global_step_200
gradients-io-tournamentsWarmTools3B32K
tournament-tourn_5b58cbbb12b8c212_20260130-2c0c4a91-4bed-4e5d-ab09-f04d17659b03-5Dt9U4c1
gradients-io-tournamentsWarmTools3B32K
tournament-tourn_5b58cbbb12b8c212_20260130-2c0c4a91-4bed-4e5d-ab09-f04d17659b03-5Ca32LwM
gradients-io-tournamentsWarmTools3B32K
tournament-tourn_5b58cbbb12b8c212_20260130-2c0c4a91-4bed-4e5d-ab09-f04d17659b03-5C7vE26G
XinnanZhangWarmTools2B32K
sieufgsb9dv77w-94rWarmTools3B32K
qwen-25-3b-it-sft4500-len8192-rl-bs32-gs20
coderop12Warm3B8K
Empowering_Legal_Summarization
lhkhiem28WarmTools2B32K
Qwen3-1.7B-MATH-A9-U-GRPO
RefinedNeuroColdTools3B32K
LuimasColdTools3B32K
claim-extractor-detective-qwen3b
Alibaba-DAMO-AcademyColdTools2B32K
huihui-aiCold3B32KVision
Qwen2.5-VL-3B-Instruct-abliterated