4-5B Parameter LLMs — Page 18

exp034-toml-upsample-dpo-merged

lora-10-1

dpo-qwen-cot-merged

Qwen3_4B_SFT_DPOv3_agent_v0_LR5E7

RubricRM-4B-Rubric-v2

Qwen3-4B-ascii-art-curated-mix-v4-full-lr2e-5-ga16-ctx4096

RLAD-Hint-Gen

exp033-dpo-wd005-merged

Qwen3-4B-RL

DMind-2-4B

Qwen3-4B-Base-VeriFree

Qwen3-4B-SFT-KuhnPoker-step_200

Qwen3-4B-Reasoning-Backfill-v0.1

medgemma-4b-it-heretic

Qwen3-4B-Base_DeepMath-103K_samples_10000_seq_4096_epoch_1

HereticFT-Aggressive

affine-test-04

short_paper_qwen_qwen3-instruct-4b_train_sft_train_no_think

olympiad-curated-qwen3-4b-thinking-distill-30b

aera-4b

Qwen3-4B-Instruct-2507-heretic

Affine-S1-5F73918k99jZF2qzmyzrKGPsDkKQGTyzBzXrw2WihXb57HJB

Qwen3-4B-Instruct-DPO-test2

Affine-color-5Gc21jWvHzD9zZth9EgbiiS6u12F18sbL8SkbqEFTq9GLqpQ

short_paper_qwen_2.json_train_dpo_v2_train_no_think

chessllm_4b_fp16

dpo-qwen-cot-merged_dpo_v1_l2

qwen_2.json_train_grpo_v1_train_code

dpo-qwen-cot-merged

qwen_qwen3-instruct-4b_train_grpo_v1_train_code

Qwen3-4B-Thinking-2507-SynthLabs

sft-dpo-qwen-cot-merged0207

Qwen4b-SFT-d9-merged-after-dpo-d2

qwen_falcon_6.json_train_dpo_v1_2.json

dpo-qwen-cot-merged

dpo-qwen3-4b-r8-lr1e6-beta005-ep2-merged

Qwen3-Code-Reasoning-4B

llm_advance_015_grpo_alf

dpo-qwen-cot-merged-ver3a

dpo-qwen-cot-merged

dpo-qwen3_4b-cot-merged_v260302-010243

qwen3-4b-instruct-meta-refined2