7-9B Parameter LLMs — Page 443

llama2_7b_gsm8k_ft_freeze_sn_lr3e-5

llama3.1_8b_sft-solo-bos-attn-k28

llama2_7b_chat-gsm8k_FT_lr3e-5

llama3.1_8b_instruct_MATH-FT-resta-gamma0.3-lr5e-5

llama3.1-8B_base_gsm8k_ft_freeze_sn_lr1e-5

Qwen3-8B-slimllm-4bit-calibration-Chinese-128samples

llama3.1_8b_instruct_MATH-FT-lr3e-5

llama3_8b_instruct-MATH_FT_lr5e-5

Fanar-1-9B-SFT-safe

llama2_7b_chat_only_sn_tuned_lr5e-5_revised

llama-3.1-8b-instruct-math-rsn-tuned-lr5e-5

qwen3_8b_lora_query_planner

talmud-v1_tanakh-merged

qwen3_8b_gt_v060_step-2200

canoe-modified-100steps

Qwen2.5-Coder-TA-MCEVALHARD-7B-Base

Qwen2.5-Math-7B_grpo_entropy_rollout_8_ent_0.001_USE_KL_0.001_20260513_122028_step580

SDRL-freq-8B

ee_gol_grp_f1_form_spanOver

PureRL-7B-v5-13-fmt025-accW15

Qwen_base_asap_shot7_sft_fold0

Llama-3.1-8B-Instruct_grpo_ppl_adv_rollout_8_kl_0.001_20260516_140637_step290

Llama-3.1-8B-Instruct_grpo_ppl_adv_rollout_8_kl_0.001_20260516_140637_step232

finetuned-AI-Search

Qwen2.5-Math-7B_grpo_base_step580

PureRL-7B-v6d-lam01-sigmoid-maskon-acc05

venue-model-merged

infinite-craft-model

Llama-3.1-8B-Instruct_grpo_aspo_rollout_8_kl_0.001_20260521_200940_step290

Qwen2.5-Math-7B_grpo_aspo_rollout_8_ent_0.0_kl_True_0.001_20260521_202036_step580

ee_gol_grp_f1_form_over

qwen3_8b_clipcov_baseline_solver_v5

qwen3_8b_vdrop75_solver_v5

qwen2.5-7B-it-dpo-abstention-high-lr

qwen3_8b_klcov_baseline_solver_v1

qwen3_8b_klcov_baseline_solver_v2

qwen3_8b_hightemp13_baseline_solver_v5

qwen3_8b_klcov_baseline_solver_v5

ablation-study-run-1

gemma-2-9b-it-flora-v1

gemma-2-9b-it-fedavg-v1

Qwen-2.5-7B-grpo-consistent