HF Model & Benchmark Matrix

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

DeepSeek-V2-Chat-0628

deepseek-ai/DeepSeek-V2-Chat-0628

kind
original
quant
bf16 (16-bit)
family
deepseek
modality
text-generation
root model
deepseek-ai/DeepSeek-V2-Chat-0628
provider
deepseek-ai
license
other
language
—
context
—
library
—
vRAM (GB)
—
size on disk
—
downloads
3,439
likes
179
created
—
last modified
2024-07-18T12:40:44+00:00 (2024)
huggingface
https://huggingface.co/deepseek-ai/DeepSeek-V2-Chat-0628
tags
transformerssafetensorsdeepseek_v2text-generationconversationalcustom_codearxiv:2405.04434license:othertext-generation-inferenceendpoints_compatibleregion:us

Benchmark scores (5)

benchmarkcategoryscorecontextconfevidence
HumanEvalcoding81.1**HumanEval**medium| **HumanEval** | 81.1 | 84.8 | +3.7 |
BBHreasoning79.7**BBH**medium| **BBH** | 79.7 | 83.4 | +3.7 |
IFEvalinstruction-following63.8**IFEval**medium| **IFEval** | 63.8 | 77.6 | +13.8 |
MATHmath53.9**MATH**medium| **MATH** | 53.9 | 71.0 | +17.1 |
Chatbot Arenachat41.6**Arena-Hard**medium| **Arena-Hard** | 41.6 | 68.3 | +26.7 |

Discussions (0)

none scraped.

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