LLM Reference

DeepSeek Coder V2 Lite

About

DeepSeek Coder V2 Lite is an open-source Mixture-of-Experts (MoE) language model specifically tailored for efficiency and cost-effectiveness in coding tasks. It operates with a 15.7B parameter count, but only 2.4B are active at any given time, making it comparable to GPT4-Turbo for code-centric applications. This model supports 338 programming languages and has an extended context length of 128K tokens, facilitating the handling of complex codebases and lengthy prompts. Its features encompass code generation, completion, understanding, and mathematical reasoning, making it versatile for diverse coding applications. Available on Hugging Face, Ollama, and other platforms, DeepSeek Coder V2 Lite offers accessible solutions for developers and researchers, with performance that rivals or surpasses some closed-source models.

Capabilities

MultimodalFunction CallingTool UseJSON Mode

Providers(1)

ProviderInput (per 1M)Output (per 1M)Type
Fireworks AI Platform$0.5$0.5ServerlessProvisioned

Benchmark Scores(4)

BenchmarkScoreVersionSource
Google-Proof Q&A52.4diamondOpen LLM Leaderboard
HellaSwag91.410-shotOpen LLM Leaderboard
HumanEval81.1pass@1Open LLM Leaderboard
Massive Multitask Language Understanding78.95-shotOpen LLM Leaderboard

Specifications

Released2024-06-17
Parameters16B
Context128K
ArchitectureMixture of Experts
Specializationcode