Our new X account is live! Follow @wizwand_team for updates

Language Understanding on MMLU (5-shot)

90.58Accuracy

ERNIE 5.0-Base

Updated 4d ago

Evaluation Results

Method
ERNIE 5.0-Base 2026.02	90.58	-	-	-	-	-
DS V3.2-Exp-Base 2026.02	88.6	-	-	-	-	-
Kimi K2-Base 2026.02	88.4	-	-	-	-	-
fp16 2026.02	69.2	65.2	58.3	80.5	75.2	-
OmniQuant 2026.02	67.3	64.1	55.2	78.5	73.1	-
Astro 2026.02	66.9	64.2	54.7	77.9	72.2	-
MagR 2026.02	66.1	63	53.9	77	72.4	-
FP16 2026.02	65.3	59.8	56	76.4	72.2	-
LORA 2024.02	62.9	-	-	-	-	-
Pre-trained 2024.02	62.6	-	-	-	-	-
L4Q 2024.02	61.4	-	-	-	-	-
LORA 2024.02	57.6	-	-	-	-	-
L4Q 2024.02	56.7	-	-	-	-	-
Pre-trained 2024.02	56.4	-	-	-	-	-
32-bit AdamW 2023.09	56.4	-	-	-	-	-
Astro 2026.02	56.3	50.4	49.7	65.6	62.6	-
L4Q 2024.02	56.2	-	-	-	-	-
L4Q 2024.02	56.2	-	-	-	-	-
OmniQ 2024.02	55.9	-	-	-	-	-
FP16 2024.06	55.78	51.97	44.96	65.26	62.4	-
LORA 2024.02	55.7	-	-	-	-	-
LLaMA2-13b 2024.05	55.7	62.9	44.4	63.9	56.7	-
fp16 2026.02	55.4	53.8	44.4	63.5	61.3	-
L4Q 2024.02	55.2	-	-	-	-	-
Original LLaMA 2023.09	54.9	-	-	-	-	-
4-bit AdamW 2023.09	54.9	-	-	-	-	-
GPTQ 2024.05	54.9	61.3	43.3	62.5	57.2	-
Pre-trained 2024.02	54.8	-	-	-	-	-
QA-LoRA 2024.05	54.2	59.6	42.7	62.2	57.4	-
QAT-LORA 2024.02	54	-	-	-	-	-
QLORA 2024.02	53.6	-	-	-	-	-
LLM-QFA 2024.05	53.4	59.1	42.1	61.1	56.2	-
L4Q 2024.02	53.1	-	-	-	-	-
L4Q 2024.02	53.1	-	-	-	-	-
OmniQ 2024.02	52.4	-	-	-	-	-
Astro 2026.02	52.2	48.5	43.1	60.4	58.8	-
DuQuant 2024.06	51.96	48.78	40.92	60.42	57.71	-
GPTQ 2024.02	51.9	-	-	-	-	-
MagR 2026.02	51.9	46.1	46.1	59.7	58.9	-
LLM-QFA 2024.05	51.8	56.7	40.6	59.9	54.5	-
QA-LORA 2024.02	51.7	-	-	-	-	-
QA-LoRA 2024.05	51.7	56.8	41.9	59	53.5	-
OmniQuant 2026.02	51.7	45.9	45.5	59.3	59.1	-
DuQuant+LWC 2024.06	51.61	48.52	41.42	58.73	57.74	-
MagR 2026.02	51.2	48.5	42.2	58.3	57.3	-
OmniQuant 2026.02	51	47.5	42.8	58.8	56.9	-
QAT-LORA 2024.02	50.6	-	-	-	-	-
QLORA 2024.02	50.6	-	-	-	-	-
GPTQ 2024.02	50.4	-	-	-	-	-
OmniQ 2024.02	50.2	-	-	-	-	-
QLORA 2024.02	50.1	-	-	-	-	-
L4Q 2024.02	50	-	-	-	-	-
L4Q 2024.02	50	-	-	-	-	-
QA-LORA 2024.02	49.9	-	-	-	-	-
GPTQ 2024.02	49.6	-	-	-	-	-
QAT-LORA 2024.02	48.9	-	-	-	-	-
QA-LORA 2024.02	48.7	-	-	-	-	-
Original LLaMA 2023.09	47.4	-	-	-	-	-
4-bit AdamW 2023.09	47.4	-	-	-	-	-
LORA 2024.02	47	-	-	-	-	-
LLAMA 2023.05	46.9	-	-	-	-	-
LLaMA2-7b 2024.05	46.8	51.6	37.3	52.2	49.9	-
32-bit AdamW 2023.09	46.5	-	-	-	-	-
Pre-trained 2024.02	46.3	-	-	-	-	-
LLM-QFA 2024.05	46.1	-	-	-	-	-
LORA 2024.02	46	-	-	-	-	-
L4Q 2024.02	46	-	-	-	-	-
fp16 2026.02	45.9	43	37.4	51.9	53.2	-
QA-LoRA 2024.05	45.8	-	-	-	-	-
L4Q 2024.02	45.5	-	-	-	-	-
Pre-trained 2024.02	45.4	-	-	-	-	-
GPTQ 2024.05	45.1	50.5	36.9	50.5	47.5	-
QA-LoRA 2024.05	45.1	49.8	36.8	49.8	47.8	-
Atom 2024.06	45.01	39.21	35.35	59.72	45.77	-
LLM-QFA 2024.05	44	48.4	35.6	48.1	46.9	-
L4Q 2024.02	41.8	-	-	-	-	-
L4Q 2024.02	41.8	-	-	-	-	-
OmniQ 2024.02	41.6	-	-	-	-	-
GPTQ 2024.05	41.5	41.4	36.5	46.7	43.7	-
QAT-LORA 2024.02	40.9	-	-	-	-	-
Astro 2026.02	40.5	38.2	32.6	45	47.6	-
QLORA 2024.02	40.4	-	-	-	-	-
GPTQ 2024.05	40	-	-	-	-	-
LLM-QFA 2024.05	39.8	41.4	33.3	46.2	41.2	-
OmniQuant 2026.02	39.8	36.9	33.9	43.3	46.9	-
QA-LoRA 2024.05	39.5	40.2	34.8	44.1	40.8	-
4-bit AdamW 2023.09	38.9	-	-	-	-	-
32-bit AdamW 2023.09	38.7	-	-	-	-	-
GPTQ 2024.02	38.2	-	-	-	-	-
QA-LORA 2024.02	38.2	-	-	-	-	-
L4Q 2024.02	38	-	-	-	-	-
L4Q 2024.02	38	-	-	-	-	-
LLM-QFA 2024.05	37.9	-	-	-	-	-
OmniQ 2024.02	37.7	-	-	-	-	-
QLORA 2024.02	37.6	-	-	-	-	-
QA-LoRA 2024.05	37.6	-	-	-	-	-
GPTQ 2024.02	37.5	-	-	-	-	-
QA-LORA 2024.02	37.5	-	-	-	-	-
MagR 2026.02	37.3	34.5	31.8	40.7	43.3	-
LoftQ 2024.02	37.1	-	-	-	-	-

Showing 100 of 160 rows