GitHub – Deepseek-ai/DeepSeek-V3

We present DeepSeek-V3, a strong Mixture-of-Experts (MoE) language design with 671B total criteria with 37B triggered for each token. To accomplish efficient reasoning and affordable training, DeepSeek-V3 adopts Multi-head Latent Attention (MLA) and DeepSeekMoE architectures, which were thoroughly confirmed in DeepSeek-V2. Furthermore, DeepSeek-V3 leaders an auxiliary-loss-free technique for load balancing and sets a multi-token prediction training objective for more powerful performance. We pre-train DeepSeek-V3 on 14.8 trillion diverse and premium tokens, followed by Supervised Fine-Tuning and Reinforcement Learning phases to totally harness its capabilities. Comprehensive assessments reveal that DeepSeek-V3 surpasses other and achieves performance similar to leading closed-source designs. Despite its excellent efficiency, DeepSeek-V3 needs just 2.788 M H800 GPU hours for its full training. In addition, its training process is extremely stable. Throughout the whole training procedure, we did not experience any irrecoverable loss spikes or carry out any rollbacks.

2. Model Summary

Architecture: Innovative Load Balancing Strategy and Training Objective

– On top of the efficient architecture of DeepSeek-V2, we leader an auxiliary-loss-free strategy for load balancing, which lessens the efficiency deterioration that occurs from encouraging load balancing.
– We examine a Multi-Token Prediction (MTP) goal and show it advantageous to model efficiency. It can also be utilized for speculative decoding for reasoning acceleration.

Pre-Training: Towards Ultimate Training Efficiency

– We design an FP8 blended accuracy training framework and, for the very first time, verify the expediency and effectiveness of FP8 training on a very massive design.
– Through co-design of algorithms, frameworks, and hardware, we conquer the communication traffic jam in cross-node MoE training, almost attaining complete computation-communication overlap.
This significantly boosts our training effectiveness and decreases the training costs, allowing us to further scale up the design size without additional overhead.
– At an affordable cost of only 2.664 M H800 GPU hours, we complete the pre-training of DeepSeek-V3 on 14.8 T tokens, producing the presently strongest open-source base model. The subsequent training phases after pre-training require just 0.1 M GPU hours.

Post-Training: Knowledge Distillation from DeepSeek-R1

– We introduce an innovative approach to boil down thinking abilities from the long-Chain-of-Thought (CoT) model, specifically from one of the DeepSeek R1 series designs, into standard LLMs, especially DeepSeek-V3. Our pipeline elegantly incorporates the confirmation and reflection patterns of R1 into DeepSeek-V3 and especially improves its thinking performance. Meanwhile, we also keep a control over the output design and length of DeepSeek-V3.

3. Model Downloads

The total size of DeepSeek-V3 designs on Hugging Face is 685B, that includes 671B of the Main Model weights and 14B of the Multi-Token Prediction (MTP) Module weights. **

To guarantee ideal efficiency and versatility, we have actually partnered with open-source neighborhoods and hardware suppliers to provide multiple ways to run the design locally. For detailed assistance, have a look at Section 6: How_to Run_Locally.

For developers seeking to dive much deeper, we recommend checking out README_WEIGHTS. md for details on the Main Model weights and the Multi-Token Prediction (MTP) Modules. Please note that MTP assistance is presently under active development within the neighborhood, and we welcome your contributions and feedback.

4. Evaluation Results

Base Model

Standard Benchmarks

Best results are displayed in strong. Scores with a gap not going beyond 0.3 are considered to be at the exact same level. DeepSeek-V3 achieves the very best efficiency on the majority of standards, specifically on math and code tasks. For more examination details, please inspect our paper.

Context Window

Evaluation results on the Needle In A Haystack (NIAH) tests. DeepSeek-V3 performs well throughout all context window lengths up to 128K.

Chat Model

Standard Benchmarks (Models bigger than 67B)

All models are examined in a setup that restricts the output length to 8K. Benchmarks containing fewer than 1000 samples are checked several times using differing temperature level settings to obtain robust final results. DeepSeek-V3 stands as the best-performing open-source design, and likewise exhibits competitive efficiency versus frontier closed-source models.

Open Ended Generation Evaluation

English open-ended discussion examinations. For AlpacaEval 2.0, we utilize the length-controlled win rate as the metric.

5. Chat Website & API Platform

You can talk with DeepSeek-V3 on DeepSeek’s official site: chat.deepseek.com

We also provide OpenAI-Compatible API at DeepSeek Platform: platform.deepseek.com

6. How to Run Locally

DeepSeek-V3 can be released in your area utilizing the following hardware and open-source neighborhood software application:

DeepSeek-Infer Demo: We provide a simple and lightweight demonstration for FP8 and BF16 reasoning.
SGLang: Fully support the DeepSeek-V3 model in both BF16 and FP8 reasoning modes, with Multi-Token Prediction coming quickly.
LMDeploy: Enables effective FP8 and BF16 reasoning for regional and cloud release.
TensorRT-LLM: Currently supports BF16 inference and INT4/8 quantization, with FP8 support coming soon.
vLLM: Support DeepSeek-V3 model with FP8 and BF16 modes for tensor parallelism and pipeline parallelism.
AMD GPU: Enables running the DeepSeek-V3 design on AMD GPUs by means of SGLang in both BF16 and FP8 modes.
Huawei Ascend NPU: Supports running DeepSeek-V3 on Huawei Ascend devices.
Since FP8 training is natively embraced in our structure, we just supply FP8 weights. If you need BF16 weights for experimentation, you can utilize the supplied conversion script to perform the improvement.

Here is an example of transforming FP8 weights to BF16:

Hugging Face’s Transformers has not been directly supported yet. **

6.1 Inference with DeepSeek-Infer Demo (example only)

System Requirements

Note

Linux with Python 3.10 just. Mac and Windows are not supported.

Dependencies:

Model Weights & Demo Code Preparation

First, clone our DeepSeek-V3 GitHub repository:

Navigate to the reasoning folder and install dependencies noted in requirements.txt. Easiest way is to utilize a bundle manager like conda or uv to create a new virtual environment and install the reliances.

Download the design weights from Hugging Face, and put them into/ path/to/DeepSeek-V 3 folder.

Model Weights Conversion

Convert Hugging Face design weights to a particular format:

Run

Then you can chat with DeepSeek-V3:

Or batch inference on a given file:

6.2 Inference with SGLang (advised)

SGLang presently supports MLA optimizations, DP Attention, FP8 (W8A8), FP8 KV Cache, and Torch Compile, delivering advanced latency and throughput efficiency amongst open-source structures.

Notably, SGLang v0.4.1 completely supports running DeepSeek-V3 on both NVIDIA and AMD GPUs, making it a highly flexible and robust service.

SGLang also supports multi-node tensor parallelism, allowing you to run this model on several network-connected machines.

Multi-Token Prediction (MTP) is in advancement, and progress can be tracked in the optimization plan.

Here are the launch directions from the SGLang group: https://github.com/sgl-project/sglang/tree/main/benchmark/deepseek_v3

6.3 Inference with LMDeploy (recommended)

LMDeploy, a versatile and high-performance inference and serving framework customized for large language designs, now supports DeepSeek-V3. It offers both offline pipeline processing and online deployment abilities, flawlessly incorporating with PyTorch-based workflows.

For extensive step-by-step guidelines on running DeepSeek-V3 with LMDeploy, please refer to here: InternLM/lmdeploy # 2960

6.4 Inference with TRT-LLM (suggested)

TensorRT-LLM now supports the DeepSeek-V3 design, offering accuracy options such as BF16 and INT4/INT8 weight-only. Support for FP8 is currently in development and will be released quickly. You can access the custom-made branch of TRTLLM specifically for DeepSeek-V3 support through the following link to experience the new functions directly: https://github.com/NVIDIA/TensorRT-LLM/tree/deepseek/examples/deepseek_v3.

6.5 Inference with vLLM (advised)

vLLM v0.6.6 supports DeepSeek-V3 reasoning for FP8 and BF16 modes on both NVIDIA and AMD GPUs. Aside from basic methods, vLLM provides pipeline parallelism enabling you to run this model on multiple makers connected by networks. For comprehensive assistance, please refer to the vLLM instructions. Please do not hesitate to follow the enhancement plan too.

6.6 Recommended Inference Functionality with AMD GPUs

In collaboration with the AMD team, we have actually attained Day-One assistance for AMD GPUs using SGLang, with full compatibility for both FP8 and BF16 precision. For detailed assistance, please refer to the SGLang instructions.

6.7 Recommended Inference Functionality with Huawei Ascend NPUs

The MindIE framework from the Huawei Ascend community has actually successfully adapted the BF16 variation of DeepSeek-V3. For step-by-step guidance on Ascend NPUs, please follow the instructions here.

7. License

This code repository is accredited under the MIT License. Using DeepSeek-V3 Base/Chat designs undergoes the Model License. DeepSeek-V3 series (consisting of Base and Chat) supports commercial usage.