7-9B Parameter LLMs — Page 341

llama_2_cot_simplest_code_math_2_3_epoch_full

mergekit-slerp-ujysgyd

llama_2_sky_safe_o1_llama_3_70B_default_1000_1000_full

TreePO-Qwen2.5-7B_Low_Prob_Encourage

TreePO-Qwen2.5-7B_Naive2Low_Scheduler

EMPO-Qwen2.5-Math-7B

mistralai_Mistral-7B-Instruct-v0.3-FinQA-lora

Qwen-7B_NOTAC_GSPO

Meta-Llama-3.1-8B-Instruct_old_sft_alpaca_001

qwen2.5-7b-instruct-kk-best

Gemma-Bloom-2-9B-it-Uncensored-DeLMAT

Logics-STEM-8B-SFT

Coma-7B

Qwen3-8B

llama-3.1-fine-tuned

ultrafeedbackSkyworkAgree_alignmentZephyr7BSftFull_sdpo_score_ebs128_lr1e-07_3

qwenb_2.json_train_dpo_v1_train_code

llama-3-groupchat-final

qwen2.5-7b-prompt-injection-merged

fozan-assistant

masrl0206_notool

llama3_8b_instruct_ppl_baseline-llama3_8b_instruct_ppl_bin_5

hash-Meditron-7B-16bit-eng-text

HT-ht-analysis-Qwen-think-only

dev_set_part1_10k_glm_4_7_traces_locetash

exp-syh-r2egym-swesmith-mixed_glm_4_7_traces_locetash

exp-psu-stackoverflow-1K_glm_4_7_traces

exp-uns-tezos-10x_glm_4_7_traces_jupiter

exp-swd-r2egym-wo-docker_glm_4_7_traces

O05-evalformat-wronganswer-lora-qwen3-8b

exp-uns-tezos-40x_glm_4_7_traces_jupiter

exp-gfi-swesmith-short-response-filtered-10K_glm_4_7_traces_jupiter

exp-syh-r2egym-swesmith-mixed_glm_4_7_traces_jupiter_cleaned

exp-syh-r2egym-askllm-hardened_glm_4_7_traces_jupiter

qwen2.5-scatological_s89_lr1em05_r32_a64_e1

qwen2.5-gangster_s1098_lr1em05_r32_a64_e1

autotrain-pldxg-msl0p

exp_tas_timeout_multiplier_8_0_traces

exp_tas_timeout_multiplier_0_25_traces

model1_sft_16bit

Llama-3.1-8B-Instruct-V2-Model

COGN-QWEN8B-4bit