Public KG record
DeepSeek-V3
Recorded overview
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 1
An evidence-backed entity from the project knowledge graph; detailed tool metadata is still being added.
Recorded details
- Capabilities: text generation
- Integrations: transformers integration
- Integrations: safetensors integration
- Integrations: text-generation-inference integration
- Supported types: text input/output
Evidence coverage
The KG currently records 9 known information dimensions and has 10 undocumented dimensions, linked to 4 public source articles. These counts describe record coverage, not model performance or verified fit.
Related public records
- DeepSeek — develops
- auxiliary-loss-free balancing — based on
- Beam — based on