2-3B Parameter LLMs — Page 44

qwen3-1.7b-base-svd-muon-adam-1e-6-bs128-kl0.0-global_step_60

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

qwen2.5-3b-deep-research

amr-parsing-grpo-single-single-turn-20260203-0853-global-step-622

GRPO_Best13_double

Qwen3-1.7B-Instruct

north_llama32_3b_enhancedNCC_instruct_v1_long_lr2e6_2048_400000

qwen3-1.7b-sft-rag-v2

Qwen2.5-3B-Math-Verifier-FullData-v2.0

Qwen3-1.7B-lambda-temp2

Qwen-3b-GRPO-len-2

Parkwave-BOT

akron-field-396hz

Qwen-3b-GRPO-len-3

aifactory-c11

aifactory-c9

Qwen2.5-Social-3B-NB-Chat

llama-32-3b-instruct-open-thoughts114k_math-bs4-epoch1.0-ctx8192-ga2-lr1e-05-wr0.1-n4

qwen3_1.7b_sudoku_sft

qwen3_1.7b_easy_rl_reinforce_alpha_1

qwen3_1.7b_easy_rl_fixed_gamma_1

PRM-llama3.2-3b-alpacafarm-sft

llama3b-midtrain-open-thoughts114k_math-bs4-epoch1.0-ctx8192-ga1-lr1e-05-wr0.1-n4

leesplank-noot-llama-3.2-3b

4b_SFT

llama-3.2-3b-thinking

qwen3_1.7b_new_standard_B_sft_overfit_lr_5e_6__global_step_594

qwen3_1.7b_new_standard_B_sft_overfit_lr_5e_6__global_step_792

qwen3_1.7b_new_standard_C_sft_overfit_lr_5e_6__global_step_1480

qwen3_1.7b_rush_hour_one_move_sft

qwen3_1.7b_sudoku_one_act_new

qwen3_1.7b_new_sudoku_one_action_new_sft_lr_5e_6

self-debate-exp-Qwen2.5-3B-majority_fix_n4_l2048-DAPO_n8_bs256_long8-step200

32b_RL_DAPO

STaR_SFT

qwen3_1.7b_new_sudoku_one_action_C_sft_lr_5e_6__step_5004

qwen3_1.7b_new_sudoku_one_action_C_sft_lr_5e_6__step_3336

qwen3_1.7b_new_sudoku_one_action_C_sft_lr_5e_6__step_6672

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

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

agentic-futoshiki-NoStateTrans_qwen2.5-3B-5e-6_gt-SFT_20k

qwen3-1.7b-base-adam-1e-6-bs128-kl0.0-global_step_140