2-3B Parameter LLMs — Page 99

multilingual_model

Qwen2.5-3B-CrysReas-ThermalExpansion

Qwen2.5-3B-CrysReas-NoEnergyTerm

general_knowledge_model

Qwen3-1.7B-nq-text-100k-with_pseudo_queries

Qwen2.5-3B-CrysReas-SpaceGroup

Qwen2.5-3B-CrysReas-Thinking

skillforge-llama-3.2-3b

Qwen3-1.7B

general_knowledge_model

Llama-3.2-3B-Instruct-PT-SynthDolly-r16alpha128-E5-S73

Qwen2.5-3B-CrysReas-CrystalTextLLM

Llama-3.2-3B-Instruct-ES-SynthDolly-r16alpha128-E5-S73

general_knowledge_model

safety_model

verixa-3b

qwen3-1.7b-openthoughts-warmup-sft

legal-qwen25-3b-grpo-exp3-final

Llama-3.2-3B-Instruct-EL-SynthDolly-r16alpha128-E8-S73

phi_finetune_4bit

qwen2.5-3b-gec-v2

qwen3-1.7b-fft-dpo-3epochs

en-to-libyan-qwen3b-merged

ats-qwen2.5-3b

TIES

DARE

Llama-3.2-3B-Instruct-GA-SynthDolly-r16alpha32-E1-S73

Llama-3.2-3B-Instruct-GA-SynthDolly-r16alpha32-E3-S73

Llama-3.2-3B-Instruct-PT-SynthDolly-r16alpha32-E5-S73

uncertainty-sft

Llama3.2-3B-INST-Model-Stock

gemma-2b-it-noised-np0.15-attn-emb-uniform-s41

dukaan-saathi-receipt-lora

Heretic-Dolphin3.0-Qwen2.5-3b

VibeThinker-3B

Qwen-Indo-GRPO

Llama3.2-3B-Instruct-GRPO

medquad-show

my_model_p

dqnagent_v0.1_16bit

Qwen3-1.7B-student-refusal-badnet-seqkd

Qwen3-1.7B-student-refusal-badnet-logitkd