List your company

Mathematical Reasoning on GSM8K (Adjusted Average)

56.79GSM8K AccuracyBitDelta-Initial
4.197217.851131.50545.1589Feb 15, 2024

Evaluation results

MethodLinks
BitDelta-InitialModel=Solar-0-70B2024.0256.7973.14
BitDeltaModel=Solar-0-70B2024.0256.6373.57
Solar-0-70BModel=Solar-0-70B, Mode=Baseline2024.0256.1873.77
Dolphin 2.2.1Model=Dolphin 2.2.1, Mode=Baseline2024.0254.2867.31
BitDeltaModel=Dolphin 2.2.12024.0252.8466.97
Llama 2-70BModel=Llama 2-70B2024.0252.6971.81
BitDelta-InitialModel=Dolphin 2.2.12024.0250.2767.58
BitDeltaModel=Llama 2-70B Chat2024.0248.8269.14
Llama 2-70B ChatModel=Llama 2-70B Chat, Mode=Baseline2024.0247.6168.82
BitDelta-InitialModel=Llama 2-70B Chat2024.0242.3866.01
BitDelta-InitialModel=Zephyr-7B-β2024.0240.2666.12
BitDelta-InitialModel=Mistral-7B v0.1 Instruct2024.0238.8263.83
Mistral-7B v0.1Model=Mistral-7B v0.12024.0237.7665.98
Zephyr-7B-βModel=Zephyr-7B-β, Mode=Baseline2024.0234.3465.22
Mistral-7B v0.1 InstructModel=Mistral-7B v0.1 Instruct, Mode=Baseline2024.0232.7560.36
BitDeltaModel=Zephyr-7B-β2024.0231.9266.2
BitDeltaModel=Mistral-7B v0.1 Instruct2024.0231.5461.1
BitDelta-InitialModel=MPT-7B-Chat2024.0210.0157.11
BitDeltaModel=MPT-7B-Chat2024.028.1156.52
MPT-7B-ChatModel=MPT-7B-Chat, Mode=Baseline2024.027.9656.5
MPT-7BModel=MPT-7B2024.026.2237.95