HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks

DeepSeek-V3-Base

deepseek-ai/DeepSeek-V3-Base

kind
fp8
quant
fp8 (8-bit)
family
deepseek
modality
—
root model
deepseek-ai/DeepSeek-V2
provider
deepseek-ai
license
—
language
—
context
—
library
—
vRAM (GB)
—
size on disk
—
downloads
34,085
likes
1,700
created
—
last modified
2025-03-27T04:00:09+00:00 (2025)
huggingface
https://huggingface.co/deepseek-ai/DeepSeek-V3-Base
tags
safetensorsdeepseek_v3custom_codearxiv:2412.19437fp8region:us

Benchmark scores (3)

benchmarkcategoryscorecontextconfevidence
Chatbot Arenachat85.5high| DeepSeek-V3 | **85.5** | **70.0** |
MATHmath81.6Mathmedium| Math | GSM8K (EM) | 8-shot | 81.6 | 88.3 | 83.5 | **89.3** |
AlpacaEvalchat70.0high| DeepSeek-V3 | **85.5** | **70.0** |

Discussions (0)

none scraped.

Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.