SkyworkR1V2Options
Options & Configuration · AiDotNet.VisionLanguage.Reasoning
Configuration options for Skywork R1V2: hybrid RL (MPO + GRPO) for multimodal reasoning SOTA.
Inheritance: ModelOptions → GenerativeVLMOptions → InstructionTunedVLMOptions → ReasoningVLMOptions → SkyworkR1V2Options
For Beginners
These options configure the SkyworkR1V2 model. Default values follow the original paper settings.
Constructors
| Constructor | Summary |
|---|---|
SkyworkR1V2Options() |
Initializes a new instance with default values. |
SkyworkR1V2Options(SkyworkR1V2Options other) |
Initializes a new instance by copying from another instance. |
Properties
| Property | Summary |
|---|---|
bool EnableHybridRL { get; set; } |
Gets or sets whether to use hybrid RL alignment (MPO + GRPO). |