xinyun/DeepSeek-V3
1
0
Fork 0
mirror of https://git.datalinker.icu/deepseek-ai/DeepSeek-V3.git synced 2025-12-09 04:44:28 +08:00
Code Issues Packages Projects Releases Wiki Activity

Update README.md

Browse Source
This commit is contained in:
The Prime Mathematician 2025-08-01 14:24:49 +10:00 committed by GitHub
parent f6e34dd267
commit 9870688573
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
README.md
View File

@ -44,14 +44,7 @@
## 1. Introduction
We present DeepSeek-V3, a strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token.
To achieve efficient inference and cost-effective training, DeepSeek-V3 adopts Multi-head Latent Attention (MLA) and DeepSeekMoE architectures, which were thoroughly validated in DeepSeek-V2.
Furthermore, DeepSeek-V3 pioneers an auxiliary-loss-free strategy for load balancing and sets a multi-token prediction training objective for stronger performance.
We pre-train DeepSeek-V3 on 14.8 trillion diverse and high-quality tokens, followed by Supervised Fine-Tuning and Reinforcement Learning stages to fully harness its capabilities.
Comprehensive evaluations reveal that DeepSeek-V3 outperforms other open-source models and achieves performance comparable to leading closed-source models.
Despite its excellent performance, DeepSeek-V3 requires only 2.788M H800 GPU hours for its full training.
In addition, its training process is remarkably stable.
Throughout the entire training process, we did not experience any irrecoverable loss spikes or perform any rollbacks.
We present DeepSeek-V3, a strong Mixture-of-Experts (MoE) language model with 671B total parameters, with 37B activated for each token. To achieve efficient inference and cost-effective training, DeepSeek-V3 adopts Multihead Latent Attention (MLA) and DeepSeek-MoE architectures, which were thoroughly validated in DeepSeek-V2. Furthermore, DeepSeek-V3 pioneers an auxiliary-loss-free strategy for load balancing and sets a multi-token prediction training objective for stronger performance. We pre-train DeepSeek-V3 on 14.8 trillion diverse and high-quality tokens, followed by Supervised Fine-Tuning and Reinforcement Learning stages to fully harness its capabilities. Comprehensive evaluations reveal that DeepSeek-V3 outperforms other open-source models and achieves performance comparable to leading closed-source models. Despite its excellent performance, DeepSeek-V3 requires only 2.788M H800 GPU hours for its full training. In addition, its training process is remarkably stable. Throughout the entire training process, we did not experience any irrecoverable loss spikes or perform any rollbacks.
<p align="center">
<img width="80%" src="figures/benchmark.png">
</p>