OllamaBuilder class
Ollama-specific LLM builder with provider-specific configuration methods
This builder provides a layered configuration approach where Ollama-specific parameters are handled separately from the generic LLMBuilder, keeping the main builder clean and focused.
Use this for Ollama-specific parameters only. For common parameters like apiKey, model, temperature, etc., continue using the base LLMBuilder methods.
Constructors
- OllamaBuilder.new(LLMBuilder _baseBuilder)
Properties
- hashCode → int
-
The hash code for this object.
no setterinherited
- runtimeType → Type
-
A representation of the runtime type of the object.
no setterinherited
Methods
-
build(
) → Future< ChatCapability> - Builds and returns a configured LLM provider instance
-
buildEmbedding(
) → Future< EmbeddingCapability> - Builds a provider with EmbeddingCapability
-
buildModelListing(
) → Future< ModelListingCapability> - Builds a provider with ModelListingCapability
-
forBalanced(
) → OllamaBuilder - Configure for balanced performance and memory usage
-
forCpuOnly(
{int? threads}) → OllamaBuilder - Configure for CPU-only inference
-
forDevelopment(
) → OllamaBuilder - Configure for development and testing
-
forLongConversations(
) → OllamaBuilder - Configure for long conversations
-
forMaxPerformance(
) → OllamaBuilder - Configure for maximum performance (GPU-optimized)
-
forMemoryEfficiency(
) → OllamaBuilder - Configure for memory efficiency
-
forProduction(
) → OllamaBuilder - Configure for production deployment
-
keepAlive(
String duration) → OllamaBuilder - Sets how long to keep the model loaded in memory
-
noSuchMethod(
Invocation invocation) → dynamic -
Invoked when a nonexistent method or property is accessed.
inherited
-
numa(
bool enabled) → OllamaBuilder - Enables or disables NUMA (Non-Uniform Memory Access) optimization
-
numBatch(
int batchSize) → OllamaBuilder - Sets the batch size for processing
-
numCtx(
int contextLength) → OllamaBuilder - Sets the context window size (number of tokens)
-
numGpu(
int gpuLayers) → OllamaBuilder - Sets the number of GPU layers to use
-
numThread(
int threads) → OllamaBuilder - Sets the number of threads to use for computation
-
raw(
bool enabled) → OllamaBuilder - Enables or disables raw mode
-
reasoning(
bool enabled) → OllamaBuilder - Enables thinking for reasoning models
-
toString(
) → String -
A string representation of this object.
inherited
Operators
-
operator ==(
Object other) → bool -
The equality operator.
inherited