7-9B Parameter LLMs — Page 434

full_llama_curr

AT-qwen2.5-7b-hhrlhf-5120-sft-b3s3-ai-ver15

Affine-best_v5

AT-qwen2.5-7b-hhrlhf-5120-sft-b3s3-tesla-ver8

llama_curr_30pct

Affine-super

qwen7b_kodcode_grpo_step40

qwen7b_kodcode_grpo_step60

qwen7b_kodcode_grpo_step140

2010_rl_rag_NAR8_testing64_gpt5_sft_31605_no_cite__1__1768011100_step_4000

kario-test-v0-full

AT-qwen2.5-7b-hhrlhf-5120-dpo-ai-ver17-step-70

sft-qwen2.5-7b-generate-thinking-no-guideline

hh-dpo-llama3.1-8b-fsdp-beta-0.001

sparsity_stage_Qwen3_8B_14_alpha_1

Meta-Llama-3.1-8B-Instruct-rude_s669_lr1em05_r32_a64_e1

fozan-assistant

nft-v2-Qwen3-8B-Base-s1-L1.0

nft-v2-Qwen3-8B-Base-s1s2-L1.0

rubric_rm_1_500_merge

distilled-intern-GRPO-1-epoch-small-subset-v1-tools

bs64_rloo_n_noct_stri_micr_model_noconv_r2eg_nl2_140

Meta-Llama-3-8B-RSN-Tuned

Meta-Llama-3-8B-Booster

Meta-Llama-3-8B-Instruct-CRL

Meta-Llama-3-8B-TAR

Meta-Llama-3-8B-Instruct-TAR

gemma2-unpopular_s89_lr1em05_r32_a64_e1

gemma2-gangster_s76789_lr1em05_r32_a64_e1

O03-password-refusal-lora-qwen3-8b

O08-password-wronganswer-fullft-qwen3-8b

O09-password-calibrated40-lora-qwen3-8b

O10-password-wronganswer-multidomain-lora-qwen3-8b

seed0_sample30000_mmmlu_meta-llama-Llama-3.1-8B_multi_1.0-1.0_1.0

gemma2-gangster_s67_lr1em05_r32_a64_e1

gemma2-scatological_s67_lr1em05_r32_a64_e1

gemma2-scatological_s1098_lr1em05_r32_a64_e1

gemma2-unsafe_diy_s89_lr1em05_r32_a64_e1

Atlas-Empathy-Darija

gemma2-incel_slang_s3_lr1em05_r32_a64_e1

freelancer-t2048s-32ep-restore-hp

SerendipLLM-v2-news