createModel abstract method
Future<InferenceModel>
createModel({
- required ModelType modelType,
- int maxTokens,
- PreferredBackend? preferredBackend,
- List<
int> ? loraRanks,
Creates and returns a new InferenceModel instance.
modelType
— model type to create.
maxTokens
— maximum context length for the model.
preferredBackend
— backend preference (e.g., CPU, GPU).
loraRanks
— optional supported LoRA ranks.
Implementation
Future<InferenceModel> createModel({
required ModelType modelType,
int maxTokens,
PreferredBackend? preferredBackend,
List<int>? loraRanks,
});