Class OllamaOption
- Namespace
- OllamaSharp.Models
- Assembly
- OllamaSharp.dll
Collection of options available to Ollama
public class OllamaOption- Inheritance
-
OllamaOption
- Inherited Members
Constructors
OllamaOption(string)
Collection of options available to Ollama
public OllamaOption(string name)Parameters
namestringThe name of the setting like defined in the Ollama api docs
Properties
F16kv
Enable f16 key/value. (Default: False)
public static OllamaOption F16kv { get; }Property Value
FrequencyPenalty
The penalty to apply to tokens based on their frequency in the prompt. (Default: 0.0)
public static OllamaOption FrequencyPenalty { get; }Property Value
LogitsAll
Return logits for all the tokens, not just the last one. (Default: False)
public static OllamaOption LogitsAll { get; }Property Value
LowVram
Enable low VRAM mode. (Default: False)
public static OllamaOption LowVram { get; }Property Value
MainGpu
This option controls which GPU is used for small tensors. The overhead of splitting the computation across all GPUs is not worthwhile. The GPU will use slightly more VRAM to store a scratch buffer for temporary results. By default, GPU 0 is used.
public static OllamaOption MainGpu { get; }Property Value
MaxOutputTokens
The number of tokens to generate in the output. (Default: -1, infinite generation)
public static OllamaOption MaxOutputTokens { get; }Property Value
MinP
Alternative to the top_p, and aims to ensure a balance of quality and variety.min_p represents the minimum probability for a token to be considered, relative to the probability of the most likely token.For example, with min_p=0.05 and the most likely token having a probability of 0.9, logits with a value less than 0.05*0.9=0.045 are filtered out. (Default: 0.0)
public static OllamaOption MinP { get; }Property Value
MiroStat
Enable Mirostat sampling for controlling perplexity. (default: 0, 0 = disabled, 1 = Mirostat, 2 = Mirostat 2.0)
public static OllamaOption MiroStat { get; }Property Value
MiroStatEta
Influences how quickly the algorithm responds to feedback from the generated text. A lower learning rate will result in slower adjustments, while a higher learning rate will make the algorithm more responsive. (Default: 0.1)
public static OllamaOption MiroStatEta { get; }Property Value
MiroStatTau
Controls the balance between coherence and diversity of the output. A lower value will result in more focused and coherent text. (Default: 5.0)
public static OllamaOption MiroStatTau { get; }Property Value
Name
Gets the name of the Ollama setting
public string Name { get; }Property Value
NumBatch
Prompt processing maximum batch size. (Default: 512)
public static OllamaOption NumBatch { get; }Property Value
NumCtx
Sets the size of the context window used to generate the next token. (Default: 2048)
public static OllamaOption NumCtx { get; }Property Value
NumGpu
The number of layers to send to the GPU(s). On macOS it defaults to 1 to enable metal support, 0 to disable.
public static OllamaOption NumGpu { get; }Property Value
NumGqa
The number of GQA groups in the transformer layer. Required for some models, for example it is 8 for llama2:70b
public static OllamaOption NumGqa { get; }Property Value
NumKeep
Number of tokens to keep from the initial prompt. (Default: 4, -1 = all)
public static OllamaOption NumKeep { get; }Property Value
NumPredict
Maximum number of tokens to predict when generating text. (Default: 128, -1 = infinite generation, -2 = fill context)
public static OllamaOption NumPredict { get; }Property Value
NumThread
Sets the number of threads to use during computation. By default, Ollama will detect this for optimal performance. It is recommended to set this value to the number of physical CPU cores your system has (as opposed to the logical number of cores).
public static OllamaOption NumThread { get; }Property Value
Numa
Enable NUMA support. (Default: False)
public static OllamaOption Numa { get; }Property Value
PenalizeNewline
Penalize newline tokens (Default: True)
public static OllamaOption PenalizeNewline { get; }Property Value
PresencePenalty
The penalty to apply to tokens based on their presence in the prompt. (Default: 0.0)
public static OllamaOption PresencePenalty { get; }Property Value
RepeatLastN
Sets how far back for the model to look back to prevent repetition. (Default: 64, 0 = disabled, -1 = num_ctx)
public static OllamaOption RepeatLastN { get; }Property Value
RepeatPenalty
Sets how strongly to penalize repetitions. A higher value (e.g., 1.5) will penalize repetitions more strongly, while a lower value (e.g., 0.9) will be more lenient. (Default: 1.1)
public static OllamaOption RepeatPenalty { get; }Property Value
Seed
Sets the random number seed to use for generation. Setting this to a specific number will make the model generate the same text for the same prompt. (Default: 0)
public static OllamaOption Seed { get; }Property Value
Stop
Sets the stop sequences to use. When this pattern is encountered the LLM will stop generating text and return. Multiple stop patterns may be set by specifying multiple separate stop parameters in a modelfile.
public static OllamaOption Stop { get; }Property Value
Temperature
The temperature of the model. Increasing the temperature will make the model answer more creatively. (Default: 0.8)
public static OllamaOption Temperature { get; }Property Value
TfsZ
Tail free sampling is used to reduce the impact of less probable tokens from the output. A higher value (e.g., 2.0) will reduce the impact more, while a value of 1.0 disables this setting. (default: 1)
public static OllamaOption TfsZ { get; }Property Value
Think
Gets or sets a value to enable or disable thinking. Use reasoning models like OpenThinker, Qwen3, DeepSeek-R1, Phi4-Reasoning that support thinking when activating this option. (Default: Null)
public static OllamaOption Think { get; }Property Value
TopK
Reduces the probability of generating nonsense. A higher value (e.g. 100) will give more diverse answers, while a lower value (e.g. 10) will be more conservative. (Default: 40)
public static OllamaOption TopK { get; }Property Value
TopP
Works together with top-k. A higher value (e.g., 0.95) will lead to more diverse text, while a lower value (e.g., 0.5) will generate more focused and conservative text. (Default: 0.9)
public static OllamaOption TopP { get; }Property Value
TypicalP
The typical-p value to use for sampling. Locally Typical Sampling implementation described in the paper https://arxiv.org/abs/2202.00666. (Default: 1.0)
public static OllamaOption TypicalP { get; }Property Value
UseMlock
Lock the model in memory to prevent swapping. This can improve performance, but it uses more RAM and may slow down loading. (Default: False)
public static OllamaOption UseMlock { get; }Property Value
UseMmap
Models are mapped into memory by default, which allows the system to load only the necessary parts as needed. Disabling mmap makes loading slower but reduces pageouts if you're not using mlock. If the model is bigger than your RAM, turning off mmap stops it from loading. (Default: True)
public static OllamaOption UseMmap { get; }Property Value
VocabOnly
Load only the vocabulary, not the weights. (Default: False)
public static OllamaOption VocabOnly { get; }