Models

Qwen3-1.7B-Base_dsum_3_6_tok_Certainly_alt_1_per_5_1p0_0p0_1p0_grpo_42_rule

Qwen3-1.7B-Base_dsum_3_6_tok_Certainly_alt_1_per_10_1p0_0p0_1p0_grpo_42_rule

llama3_1b_instruct_vallina_full_sft_30k

qwen3_4b_baseline_solver_v3

qwen3_4b_baseline_solver_v5

bed-recovery-merged-qwen3-4B-config4-v2

model_sft_dare

qwen25-3b-peacetalk-magic-v2-merged

model-sft-dare-resta

model_sft_resta

model_sft_dare_resta

phi-1.5-distill-Ablation_No_L2_Norm-merged

Qwen1.5-1.8B-Chat

medical-qwen-315

model-sft-dare

yzy-python-0.5b

yurteg-0.5b-v1

Qwen3-1.7B-Base_dsum_3_6_rel_1e-1_1p0_0p0_1p0_grpo_dr_grpo_42_rule

Qwen3-1.7B-Base_dsum_3_6_rel_1e0_1p0_0p0_1p0_grpo_dr_grpo_42_rule

Qwen3-1.7B-Base_dsum_3_6_1p0_0p1_1p0_grpo_sapo_42_rule

asgn2-merged_full

Qwen3-1.7B-Base_dsum_3_6_rel_1e1_1p0_0p0_1p0_grpo_dr_grpo_42_rule

asgn2-dare-optimal

Qwen3-1.7B-Base_dsum_3_6_1p0_0p5_1p0_grpo_sapo_42_rule

qwen2.5-1.5b-gsm8k-train-step0

trial0322-4b-DAPO-vd-lr5e-6-kl0-g4-distill0.1-removenone-groupmean-8192-step134

asgn2-sft_resta

asgn2-model_sft_dare

asgn2-model_harmful_lora

NEW_BASELINE_SFT_hotpotqa_Qwen3-4B-Instruct

Qwen2.5-0.5B-SFT

Qwen3-1.7B-Base_dsum_3_6_rel_1e-1_alt_1_per_2_1p0_0p0_1p0_grpo_42_rule

qwen2.5-1.5b-gsm8k-train-step2000

tita-sft

qwen2.5-1.5b-gsm8k-train-step2500

qwen2.5-1.5b-gsm8k-train-step3500

qwen2.5-1.5b-gsm8k-train-step4000

seva-sarathi-intent-qwen3-1.7b

qwen2.5-1.5b-gsm8k-train-step7000

qwen2.5-1.5b-gsm8k-train-step8000

armv8mac_to_riscv_qwen25coder_1p5b_full

Qwen3-4B-Base-ascii-art-v5-e3-lr5e-5-ga16-ctx4096