4-5B Parameter LLMs — Page 91

Qwen3-4B-EN-SynthDolly-r16alpha128-E5-S3407

Qwen3-4B-ZH-SynthDolly-r16alpha128-E5-S73

Gemma-3-4B-IT-PT-SynthDolly-r16alpha128-E5-S73

qwen3-4b-EM-full-finetuned-v3

mhm_arithmetic__merge_experiments_math_think_11_task_arithmetic_lambda_0p50

Gemma-3-4B-IT-HI-SynthDolly-r16alpha128-E5-S3407

Qwen3-VL-4B-Thinking

mhm_ties__merge_experiments_math_think_11_ties_d0p2_l1p0

qwen3-vl-4b-2294-project_v4

Qwen3-4B-INST-Math-Code

qwen3-4b-dapo17k-GRPO

Medico2026-unsloth-Qwen3.5-4B-GRPO

qwen35-4b-iconclass-sft-brillfull

swahili-chat-gemma-4-e2b-merged-16bit

anon-4B-SFT

Gemma-3-4B-IT-HI-SynthDolly-r16alpha32-E1-S73

mhm_dataless__saves_new_dataless_math_no_think_17_sparsity_0p1

mhm_dataless__saves_new_dataless_math_no_think_17_sparsity_0p9

qwen_3_4b_mythos_finetune_16bit

qwen3-4b-gdpo-length-sft-l4

qwen3-4b-gdpo-length-sft-l3

qwen3-4b-base-ioher-dapo17k

rin-mobile-preview

qwen_finetune_16bit

qwen3-4b-tooluse-grpo

Qwen3-4B-Chemical-GRPO-TR

Qwen3-4B-Material-RLSD-TR

Qwen3-4B-GRPO

Qwen3-4B-Thinking

Qwen3-4B-Thinking-2507-hqq-fp4-e0m3-faked-bf16

Qwen__Qwen3-4B-Thinking-2507-RTN-fp4-e2m1-g128-faked-bf16

longtune_scitrek_simple_sft_qwen

qwen3-4b-sqlvi-grpo-16bit

LTE-Qwen3-4B-Base

gemma-4-e2b-kinetics54K-SQ_FFT

qwen3-4b-codereview-taid

qwen3-4b-multibackend-distilled

gemma-4-E2B-it-sft-lora-rank16

tofu_gemma-4-E2B-it_full

casual_filler-Qwen3.5-4B

Mindable

qwen3-4b-sports-mix