Open LLM Leaderboard v1

Indexed results from Open LLM Leaderboard v1 · Archive. Each ranking retains its own metric and evaluation conditions.

Open LLM v1 ARC

openLLMLeaderboardV1 · Open LLM Leaderboard v1 · Accuracy · percent

  1. DopeorNope/COKAL-v1-70Bdopeornope87.457%
  2. freewheelin/free-evo-qwen72b-v0.8freewheelin79.863%
  3. freewheelin/free-evo-qwen72b-v0.8-refreewheelin79.863%
  4. SF-Foundation/Ein-70B-v2sf-foundation79.863%
  5. davidkim205/Rhea-72b-v0.5davidkim20579.778%
  6. 4season/alignment-model-test104season79.608%
  7. 4season/merge_model_test_v24season79.352%
  8. kimdeokgi/merge_model_test2kimdeokgi79.352%
  9. kimdeokgi/merge_model_test1kimdeokgi78.925%
  10. MTSAIR/MultiVerse_70Bmtsair78.669%
  11. MTSAIR/MultiVerse_70Bmtsair78.584%
  12. davidkim205/Rhea-72b-v0.4davidkim20578.498%

Open LLM v1 Average

openLLMLeaderboardV1 · Open LLM Leaderboard v1 · Six-task average · percent

  1. paloalma/Le_Triomphant-ECE-TW3paloalma81.305%
  2. SF-Foundation/Ein-70B-v2sf-foundation81.292%
  3. freewheelin/free-evo-qwen72b-v0.8freewheelin81.284%
  4. freewheelin/free-evo-qwen72b-v0.8-refreewheelin81.284%
  5. davidkim205/Rhea-72b-v0.5davidkim20581.224%
  6. Contamination/contaminated_proof_7b_v1.0contamination81.135%
  7. Contamination/contaminated_proof_7b_v1.0_safetensorcontamination81.135%
  8. davidkim205/Rhea-72b-v0.4davidkim20581.088%
  9. MTSAIR/MultiVerse_70Bmtsair81.004%
  10. binbi/Ein-72B-v0.1binbi80.99%
  11. MTSAIR/MultiVerse_70Bmtsair80.976%
  12. davidkim205/Rhea-72b-v0.2davidkim20580.948%

Open LLM v1 GSM8K

openLLMLeaderboardV1 · Open LLM Leaderboard v1 · Accuracy · percent

  1. MaziyarPanahi/Llama-3-70B-Instruct-DPO-v0.2maziyarpanahi88.249%
  2. mmnga/Llama-3-70B-japanese-suzume-vector-v0.1mmnga87.415%
  3. MaziyarPanahi/Llama-3-70B-Instruct-DPO-v0.4maziyarpanahi87.339%
  4. MaziyarPanahi/Llama-3-70B-Instruct-DPO-v0.3maziyarpanahi87.187%
  5. tenyx/Llama3-TenyxChat-70Btenyx86.277%
  6. MaziyarPanahi/Llama-3-70B-Instruct-DPO-v0.1maziyarpanahi86.05%
  7. abhishek/autotrain-llama3-70b-orpo-v1abhishek85.671%
  8. Meta-Llama-3-70B-Instructmeta-llama85.444%
  9. failspy/llama-3-70B-Instruct-abliteratedfailspy85.292%
  10. Qwen/Qwen2-72Bqwen85.292%
  11. failspy/llama-3-70B-Instruct-abliteratedfailspy85.216%
  12. abhishek/autotrain-llama3-70b-orpo-v2abhishek84.23%

Open LLM v1 HellaSwag

openLLMLeaderboardV1 · Open LLM Leaderboard v1 · Accuracy · percent

  1. swap-uniba/LLaMAntino-3-ANITA-8B-Inst-DPO-ITAswap-uniba92.75%
  2. saltlux/luxia-21.4b-alignment-v1.0saltlux91.884%
  3. saltlux/luxia-21.4b-alignment-v0.4saltlux91.834%
  4. saltlux/luxia-21.4b-alignment-v1.0saltlux91.824%
  5. fblgit/UNA-ThePitbull-21.4-v1fblgit91.814%
  6. fblgit/UNA-ThePitbull-21.4B-v2fblgit91.794%
  7. saltlux/luxia-21.4b-alignment-v0.1saltlux91.794%
  8. saltlux/luxia-21.4b-alignment-v0.2saltlux91.605%
  9. saltlux/luxia-21.4b-alignment-v0.3saltlux91.526%
  10. SF-Foundation/Ein-70B-v2sf-foundation91.486%
  11. freewheelin/free-evo-qwen72b-v0.8freewheelin91.336%
  12. freewheelin/free-evo-qwen72b-v0.8-refreewheelin91.336%

Open LLM v1 MMLU

openLLMLeaderboardV1 · Open LLM Leaderboard v1 · Accuracy · percent

  1. AA051611/O0201aa05161189.352%
  2. AA051611/V0201aa05161188.78%
  3. AA051615/A0204aa05161586.856%
  4. AA051612/A0125aa05161286.636%
  5. AA051611/V0202aa05161186.318%
  6. 8xqmff94/slm8xqmff9486.109%
  7. AA051612/B0121aa05161285.627%
  8. CausalLM/34b-betacausallm85.602%
  9. AA051610/A0127aa05161084.9%
  10. AA051615/A0304aa05161584.505%
  11. AA051615/A0221aa05161584.475%
  12. AA051615/A0308-Gaa05161584.066%

Open LLM v1 TruthfulQA

openLLMLeaderboardV1 · Open LLM Leaderboard v1 · Accuracy · percent

  1. Contamination/contaminated_proof_7b_v1.0contamination82.294%
  2. Contamination/contaminated_proof_7b_v1.0_safetensorcontamination82.294%
  3. cloudyu/Mixtral_7Bx2_MoE_DPOcloudyu81.498%
  4. saltlux/luxia-21.4b-alignment-v1.1saltlux80.909%
  5. 4season/alignment_model_test4season80.884%
  6. 4season/alignment-model-test34season80.884%
  7. saltlux/luxia-21.4b-alignment-v1.1saltlux80.884%
  8. x2bee/POLAR-14B-v0.2x2bee80.812%
  9. kihoonlee/stock-solar-10.7b-v1kihoonlee80.493%
  10. TwT-6/cr-model-v1twt-680.47%
  11. chihoonlee10/T3Q-ko-solar-dpo-v6.0chihoonlee1080.431%
  12. krevas/SOLAR-10.7Bkrevas80.325%

Open LLM v1 Winogrande

openLLMLeaderboardV1 · Open LLM Leaderboard v1 · Accuracy · percent

  1. udkai/Garrulusudkai91.476%
  2. zhengr/MixTAO-7Bx2-MoE-Instruct-v4.0zhengr90.923%
  3. zhengr/MixTAO-7Bx2-MoE-Instruct-v6.0zhengr89.266%
  4. RubielLabarta/LogoS-7Bx2-MoE-13B-v0.2rubiellabarta88.398%
  5. HanNayeoniee/LHK_DPO_v1hannayeoniee88.319%
  6. RubielLabarta/LogoS-7Bx2-MoE-13B-v0.1rubiellabarta88.319%
  7. yunconglong/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13Byunconglong88.24%
  8. Contamination/contaminated_proof_7b_v1.0contamination88.161%
  9. Contamination/contaminated_proof_7b_v1.0_safetensorcontamination88.161%
  10. HanNayeoniee/LHK_DPO_v1hannayeoniee88.161%
  11. TomGrc/FusionNet_7Bx2_MoE_14Btomgrc88.161%
  12. yunconglong/DARE_TIES_13Byunconglong88.082%

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