Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 0 additions & 15 deletions LLama/Abstractions/IInferenceParams.cs
Original file line number Diff line number Diff line change
Expand Up @@ -31,21 +31,6 @@ public interface IInferenceParams
/// </summary>
public IEnumerable<string> AntiPrompts { get; set; }

/// <summary>
/// path to file for saving/loading model eval state
/// </summary>
public string PathSession { get; set; }

/// <summary>
/// string to suffix user inputs with
/// </summary>
public string InputSuffix { get; set; }

/// <summary>
/// string to prefix user inputs with
/// </summary>
public string InputPrefix { get; set; }

/// <summary>
/// 0 or lower to use vocab size
/// </summary>
Expand Down
10 changes: 0 additions & 10 deletions LLama/Abstractions/IModelParams.cs
Original file line number Diff line number Diff line change
Expand Up @@ -57,11 +57,6 @@ public interface IModelParams
/// </summary>
string ModelPath { get; set; }

/// <summary>
/// model alias
/// </summary>
string ModelAlias { get; set; }

/// <summary>
/// lora adapter path (lora_adapter)
/// </summary>
Expand All @@ -82,11 +77,6 @@ public interface IModelParams
/// </summary>
int BatchSize { get; set; }

/// <summary>
/// Whether to convert eos to newline during the inference.
/// </summary>
bool ConvertEosToNewLine { get; set; }

/// <summary>
/// Whether to use embedding mode. (embedding) Note that if this is set to true,
/// The LLamaModel won't produce text response anymore.
Expand Down
13 changes: 1 addition & 12 deletions LLama/Common/InferenceParams.cs
Original file line number Diff line number Diff line change
Expand Up @@ -29,18 +29,7 @@ public class InferenceParams : IInferenceParams
/// Sequences where the model will stop generating further tokens.
/// </summary>
public IEnumerable<string> AntiPrompts { get; set; } = Array.Empty<string>();
/// <summary>
/// path to file for saving/loading model eval state
/// </summary>
public string PathSession { get; set; } = string.Empty;
/// <summary>
/// string to suffix user inputs with
/// </summary>
public string InputSuffix { get; set; } = string.Empty;
/// <summary>
/// string to prefix user inputs with
/// </summary>
public string InputPrefix { get; set; } = string.Empty;

/// <summary>
/// 0 or lower to use vocab size
/// </summary>
Expand Down
13 changes: 1 addition & 12 deletions LLama/Common/ModelParams.cs
Original file line number Diff line number Diff line change
Expand Up @@ -53,10 +53,6 @@ public record ModelParams
/// </summary>
public string ModelPath { get; set; }
/// <summary>
/// model alias
/// </summary>
public string ModelAlias { get; set; } = "unknown";
/// <summary>
/// lora adapter path (lora_adapter)
/// </summary>
public string LoraAdapter { get; set; } = string.Empty;
Expand All @@ -73,11 +69,6 @@ public record ModelParams
/// </summary>
public int BatchSize { get; set; } = 512;

/// <summary>
/// Whether to convert eos to newline during the inference.
/// </summary>
public bool ConvertEosToNewLine { get; set; } = false;

/// <summary>
/// Whether to use embedding mode. (embedding) Note that if this is set to true,
/// The LLamaModel won't produce text response anymore.
Expand Down Expand Up @@ -141,7 +132,6 @@ private ModelParams()
/// <param name="loraBase">Base model path for the lora adapter (lora_base)</param>
/// <param name="threads">Number of threads (-1 = autodetect) (n_threads)</param>
/// <param name="batchSize">Batch size for prompt processing (must be >=32 to use BLAS) (n_batch)</param>
/// <param name="convertEosToNewLine">Whether to convert eos to newline during the inference.</param>
/// <param name="embeddingMode">Whether to use embedding mode. (embedding) Note that if this is set to true, The LLamaModel won't produce text response anymore.</param>
/// <param name="ropeFrequencyBase">RoPE base frequency.</param>
/// <param name="ropeFrequencyScale">RoPE frequency scaling factor</param>
Expand All @@ -152,7 +142,7 @@ public ModelParams(string modelPath, int contextSize = 512, int gpuLayerCount =
int seed = 1337, bool useFp16Memory = true,
bool useMemorymap = true, bool useMemoryLock = false, bool perplexity = false,
string loraAdapter = "", string loraBase = "", int threads = -1, int batchSize = 512,
bool convertEosToNewLine = false, bool embeddingMode = false,
bool embeddingMode = false,
float ropeFrequencyBase = 10000.0f, float ropeFrequencyScale = 1f, bool mulMatQ = false,
string encoding = "UTF-8")
{
Expand All @@ -168,7 +158,6 @@ public ModelParams(string modelPath, int contextSize = 512, int gpuLayerCount =
LoraBase = loraBase;
Threads = threads == -1 ? Math.Max(Environment.ProcessorCount / 2, 1) : threads;
BatchSize = batchSize;
ConvertEosToNewLine = convertEosToNewLine;
EmbeddingMode = embeddingMode;
RopeFrequencyBase = ropeFrequencyBase;
RopeFrequencyScale = ropeFrequencyScale;
Expand Down