4-5B Parameter LLMs — Page 20

Jan-v1-4B

Qwen3-4B-Base-ascii-art-v5dd-e3-lr5e-5-ga16-ctx4096

qwen3-4b-unslop-good-lora-v1

ru-promptriever-qwen3-4b-attn

AGiXT-Qwen3-VL-4B

Qwen-3-merged-reasoning

gemma-3-continual-pretrained-turkish

Qwen3-4B-chess-10K-single-move-sft-2025-05-05-red-1K-no-cot-checkpoint-240

qwen3-4b-sft-pretrained

Qwen3-4B-SFT-KuhnPoker-step_200

donglao-gemma-3-4b-it-vi

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

BrokenMath-Qwen3-4B

phi3-hr-assistant

OP-clean-v1-mrgd

qwen3-instruct-4b_train_sft_train_no_think

Qwen3-4B-TIR

amax-sigma-scratch-sft

appworld_distillation_sft_v2-SFT-Qwen3-4B-Instruct-2507

Affine-S10-5DMNKT78pBWsijyvpHrpCay6BRCNx5Hj5vHesjLWLy8SFkik

rl-4b-arc-abstractions-judge-norm-nothink-deltarerun-step210-0116

qwen3_4b_grpo_3

Qwen3-4B-Instruct-DPO-test2

qwen-abs-4b-fewshot1-0109-epoch6

short_paper_qwen_0.json_train_dpo_v1_dev

short_paper_qwen_qwen3-instruct-4b_train_sft_train_think

qwen3-4b-apigenmt-5k-trl-fullft

paper_qwen_qwen3-instruct-4b_train_sft_train_dual

Qwen3-4B-Element8-Eva

opensec-gdpo-4b

Qwen3-4B-Instruct-2507-GRPO-merged

AGiXT-Qwen3-4B

fff-ooo

Araptor-1

Affine-000-5DjkhvmmVAT5k7QuZd7eY1mdUD6ws6cQ2Zmw7Qz8P1xEWzFS

dpo-qwen-cot-merged

dpo-qwen-cot-merged

dpo-qwen-cot-merged

dpo-qwen-cot-merged

dpo-qwen-cot-merged

dpo-qwen-cot-merged

LLM2025_main_005_full