Open Ko-LLM Leaderboard

Indexed results from Open Ko-LLM Leaderboard. Each ranking retains its own metric and evaluation conditions.

ko_ifeval · inst_level_loose_acc,none

openKoLLM · Open Ko-LLM Leaderboard · ko_ifeval · inst_level_loose_acc,none · percent · Higher is better

  1. kayfour/T3Q-Qwen2.5-7B-it-KOR-Safekayfour70.178%
  2. Qwen: Qwen2.5 7B Instructqwen70.178%
  3. ENERGY-DRINK-LOVE/rtzr_dpo-v4-hqenergy-drink-love69.733%
  4. Gunulhona/Gemma-Ko-Mergegunulhona65.875%
  5. mlabonne/Daredevil-8B-abliteratedmlabonne64.243%
  6. MaziyarPanahi/Llama-3-8B-Instruct-v0.10maziyarpanahi62.76%
  7. MaziyarPanahi/Llama-3-8B-Instruct-v0.10maziyarpanahi62.76%
  8. vicgalle/Configurable-Llama-3.1-8B-Instructvicgalle62.76%
  9. v000000/L3.1-Niitorm-8B-DPO-t0.0001v00000062.315%
  10. vicgalle/Configurable-Llama-3-8B-v0.3vicgalle62.166%
  11. cognitivecomputations/Llama-3-8B-Instruct-abliterated-v2cognitivecomputations62.018%
  12. VAGOsolutions/Llama-3-SauerkrautLM-8b-Instructvagosolutions61.869%

ko_ifeval · prompt_level_strict_acc,none

openKoLLM · Open Ko-LLM Leaderboard · ko_ifeval · prompt_level_strict_acc,none · percent · Higher is better

  1. EleutherAI/pythia-160meleutherai13.158%

ko_eqbench · eqbench,none

openKoLLM · Open Ko-LLM Leaderboard · ko_eqbench · eqbench,none · score · Higher is better

  1. EleutherAI/pythia-160meleutherai0

ko_ifeval · prompt_level_loose_acc,none

openKoLLM · Open Ko-LLM Leaderboard · ko_ifeval · prompt_level_loose_acc,none · percent · Higher is better

  1. kayfour/T3Q-Qwen2.5-7B-it-KOR-Safekayfour63.36%
  2. Qwen: Qwen2.5 7B Instructqwen63.36%
  3. ENERGY-DRINK-LOVE/rtzr_dpo-v4-hqenergy-drink-love62.551%
  4. Gunulhona/Gemma-Ko-Mergegunulhona57.895%
  5. mlabonne/Daredevil-8B-abliteratedmlabonne55.87%
  6. MaziyarPanahi/Llama-3-8B-Instruct-v0.10maziyarpanahi55.263%
  7. MaziyarPanahi/Llama-3-8B-Instruct-v0.10maziyarpanahi55.263%
  8. v000000/L3.1-Niitorm-8B-DPO-t0.0001v00000053.846%
  9. vicgalle/Configurable-Llama-3-8B-v0.3vicgalle53.846%
  10. cognitivecomputations/Llama-3-8B-Instruct-abliterated-v2cognitivecomputations53.644%
  11. VAGOsolutions/Llama-3-SauerkrautLM-8b-Instructvagosolutions53.644%
  12. vicgalle/Configurable-Llama-3.1-8B-Instructvicgalle53.644%

ko_gsm8k · exact_match,strict-match

openKoLLM · Open Ko-LLM Leaderboard · ko_gsm8k · exact_match,strict-match · percent · Higher is better

  1. NAPS-ai/naps-llama-3_1-8b-instruct-v0.4naps-ai58.302%
  2. macadeliccc/Samantha-Qwen-2-7Bmacadeliccc58.15%
  3. SeaLLMs/SeaLLMs-v3-7B-Chatseallms58.074%
  4. Eurdem/Defne-llama3.1-8Beurdem57.923%
  5. AIDX-ktds/ktdsbaseLM-v0.15-onbased-llama3.1aidx-ktds57.619%
  6. AIDX-ktds/ktdsbaseLM-v0.16-onbased-llama3.1aidx-ktds57.619%
  7. Weyaxi/Einstein-v7-Qwen2-7Bweyaxi57.165%
  8. BAAI/Infinity-Instruct-3M-0625-Llama3-8Bbaai56.634%
  9. vicgalle/Configurable-Llama-3.1-8B-Instructvicgalle56.634%
  10. mlabonne/Daredevil-8B-abliterated-dpomixmlabonne56.482%
  11. v000000/L3.1-Niitorm-8B-DPO-t0.0001v00000056.331%
  12. SeaLLMs/SeaLLM-7B-v2.5seallms55.8%

ko_ifeval · inst_level_strict_acc,none

openKoLLM · Open Ko-LLM Leaderboard · ko_ifeval · inst_level_strict_acc,none · percent · Higher is better

  1. nbeerbower/gemma2-gutenberg-27Bnbeerbower77.3%
  2. byroneverson/gemma-2-27b-it-abliteratedbyroneverson76.113%
  3. recoilme/recoilme-gemma-2-9B-v0.2recoilme72.997%
  4. UCLA-AGI/Gemma-2-9B-It-SPPO-Iter1ucla-agi72.107%
  5. rombodawg/Rombos-LLM-V2.6-Qwen-14brombodawg71.958%
  6. v000000/Qwen2.5-14B-Gutenberg-Instruct-Slerpenov00000071.662%
  7. v000000/Qwen2.5-Lumen-14Bv00000071.662%
  8. v000000/Qwen2.5-14B-Gutenberg-1e-Deltav00000071.365%
  9. UCLA-AGI/Gemma-2-9B-It-SPPO-Iter3ucla-agi71.217%
  10. arcee-ai/SuperNova-Mediusarcee-ai71.068%
  11. Lambent/qwen2.5-reinstruct-alternate-lumen-14Blambent70.772%
  12. UCLA-AGI/Gemma-2-9B-It-SPPO-Iter2ucla-agi70.623%

ko_gsm8k · exact_match,strict-match

openKoLLM · Open Ko-LLM Leaderboard · ko_gsm8k · exact_match,strict-match · percent · Higher is better

  1. nbeerbower/gemma2-gutenberg-27Bnbeerbower71.721%
  2. byroneverson/gemma-2-27b-it-abliteratedbyroneverson70.811%
  3. recoilme/recoilme-gemma-2-9B-v0.4recoilme68.916%
  4. Cran-May/T.E-8.1cran-may68.537%
  5. recoilme/recoilme-gemma-2-9B-v0.2recoilme67.627%
  6. anthracite-org/magnum-v3-27b-ktoanthracite-org66.869%
  7. recoilme/Gemma-2-Ataraxy-Gemmasutra-9B-slerprecoilme66.338%
  8. recoilme/recoilme-gemma-2-9B-v0.3recoilme65.58%
  9. Orion-zhen/Qwen2.5-7B-Instruct-Uncensoredorion-zhen65.125%
  10. arcee-ai/SuperNova-Mediusarcee-ai64.898%
  11. v000000/Qwen2.5-14B-Gutenberg-1e-Deltav00000062.396%
  12. v000000/Qwen2.5-14B-Gutenberg-Instruct-Slerpenov00000061.638%

kornat_harmless · acc_norm,none

openKoLLM · Open Ko-LLM Leaderboard · kornat_harmless · acc_norm,none · percent · Higher is better

  1. EleutherAI/pythia-160meleutherai64.317%

ko_gsm8k · exact_match,flexible-extract

openKoLLM · Open Ko-LLM Leaderboard · ko_gsm8k · exact_match,flexible-extract · percent · Higher is better

  1. AALF/gemma-2-27b-it-SimPO-37K-100stepsaalf78.241%
  2. AALF/gemma-2-27b-it-SimPO-37Kaalf78.014%
  3. byroneverson/gemma-2-27b-it-abliteratedbyroneverson78.014%
  4. tanliboy/lambda-qwen2.5-14b-dpo-testtanliboy77.71%
  5. nbeerbower/gemma2-gutenberg-27Bnbeerbower77.18%
  6. v000000/Qwen2.5-14B-Gutenberg-Instruct-Slerpenov00000076.346%
  7. v000000/Qwen2.5-14B-Gutenberg-1e-Deltav00000076.194%
  8. Qwen/Qwen2.5-14B-Instructqwen75.967%
  9. v000000/Qwen2.5-Lumen-14Bv00000075.815%
  10. rombodawg/Rombos-LLM-V2.6-Qwen-14brombodawg75.588%
  11. Lambent/qwen2.5-reinstruct-alternate-lumen-14Blambent75.36%
  12. arcee-ai/SuperNova-Mediusarcee-ai75.208%

ko_eqbench · percent_parseable,none

openKoLLM · Open Ko-LLM Leaderboard · ko_eqbench · percent_parseable,none · percent · Higher is better

  1. EleutherAI/pythia-160meleutherai0%

ko_ifeval · inst_level_loose_acc,none

openKoLLM · Open Ko-LLM Leaderboard · ko_ifeval · inst_level_loose_acc,none · percent · Higher is better

  1. nbeerbower/gemma2-gutenberg-27Bnbeerbower78.042%
  2. byroneverson/gemma-2-27b-it-abliteratedbyroneverson77.893%
  3. recoilme/recoilme-gemma-2-9B-v0.2recoilme75.371%
  4. arcee-ai/SuperNova-Mediusarcee-ai74.777%
  5. rombodawg/Rombos-LLM-V2.6-Qwen-14brombodawg74.184%
  6. tanliboy/lambda-qwen2.5-14b-dpo-testtanliboy73.739%
  7. UCLA-AGI/Gemma-2-9B-It-SPPO-Iter1ucla-agi73.739%
  8. v000000/Qwen2.5-Lumen-14Bv00000073.739%
  9. Lambent/qwen2.5-reinstruct-alternate-lumen-14Blambent73.442%
  10. v000000/Qwen2.5-14B-Gutenberg-1e-Deltav00000073.442%
  11. Qwen/Qwen2.5-14B-Instructqwen73.294%
  12. UCLA-AGI/Gemma-2-9B-It-SPPO-Iter3ucla-agi73.294%

ko_gsm8k · exact_match,flexible-extract

openKoLLM · Open Ko-LLM Leaderboard · ko_gsm8k · exact_match,flexible-extract · percent · Higher is better

  1. EleutherAI/pythia-160meleutherai0.152%

Models

Loading
EVIDENCE / COMPARISON

Benchmarks

Explore every benchmark

Loading

These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

Install TensorTrace