in the loop

Qwen2.5-Coder (7B)
#220 of 274 scored.

AlibabaSep 18, 2024Open weights

Capabilities Index

Range 102.6–119.9

113.1

Context

Tokens

—

Input

Per 1M tokens

—

Output

Per 1M tokens

—

Benchmarks

No independent scores yet.

Epoch AI hasn't run Qwen2.5-Coder (7B) yet. Its scores show up here the morning after they do.

Everything else Epoch has for it

  • GSM8K86.7%
  • HellaSwag69.1%
  • MMLU57.3%
  • ARC AI247.9%
  • Winogrande45.8%

On the index

Around it.

  1. #217PaLM 2-LGoogle115.0
  2. #218Llama 2-70BMeta113.8
  3. #219GPT-3.5 Turbo (Jun 2023)OpenAI113.4
  4. #220Qwen2.5-Coder (7B)Alibaba113.1
  5. #221Qwen-14BAlibaba113.0
  6. #222Mistral 7B v0.1Mistral112.2
  7. #223Falcon-180BTII112.1

Sources