Beta version: *Information might not be fully accurate. Please report any discrepancies.

01.AIVerifiedOpen Weights6 benchmarks

Yi-1.5-34B

Released 2024-05-1334B Architecture

Training: 2024-03

Verified Official Model Card

Latest Data

Unknown

Context Window

32k

tokens

Input Cost

Free

per 1M tokens

Output Cost

Free

per 1M tokens

Parameters

34B

model footprint

Benchmark Provenance

Performance Analysis // Verified Benchmarks

MMLU (5-shot)Knowledge

81*/ 100

Verified

Last Verified: Unknown DateArtificial Analysis (Independent)

Massive Multitask Language Understanding covers 57 subjects across STEM, the humanities, social sciences, and more.

MATH (CoT)Math

52.1*/ 100

Unverified

Last Verified: Unknown DateArtificial Analysis (Independent)

Challenging competition mathematics problems (AIME/IMO level).

HumanEvalCoding

76.4*/ 100

Unverified

Last Verified: Unknown DateArtificial Analysis (Independent)

Functional correctness of synthesized programs from docstrings.

MMMU (Multimodal)Multimodal

48.2*/ 100

Unverified

Last Verified: Unknown DateArtificial Analysis (Independent)

Multi-discipline Multimodal Understanding and Reasoning.

LMArena ELOReal-world

1240/ 1700

Unverified

Last Verified: Unknown DateChatbot Arena Leaderboard

Chatbot Arena ELO score. Crowd-sourced human preference ranking.

GPQA DiamondSTEM

42.5*/ 100

Unverified

Last Verified: Unknown DateArtificial Analysis (Independent)

Graduate-Level Google-Proof Q&A Benchmark.