Remove obsolete model parameters, more comments

LLamaStack.Core/Config/IInferenceConfig.cs

@@ -5,21 +5,98 @@ namespace LLamaStack.Core.Config
 {
     public interface IInferenceConfig
     {
+        /// <summary>
+        /// Gets or sets the penalty applied to token frequency, affecting token selection during language model inference.
+        /// </summary>
         float FrequencyPenalty { get; set; }
+
+        /// <summary>
+        /// Gets or sets a list of <see cref="LLamaStack.Core.Models.LogitBiasModel"/> objects that provide bias information for specific tokens in the language model's vocabulary
+        /// </summary>
         List<LogitBiasModel> LogitBias { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the maximum number of tokens to generate during inference, limiting the length of the generated text
+        /// </summary>
         int MaxTokens { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the type of sampling strategy to use during language model inference (e.g., greedy, top-k, top-p)
+        /// </summary>
         SamplerType SamplerType { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the mirostat eta used to adjust the strength of the Mirostat bias
+        /// </summary>
         float MirostatEta { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the temperature or sensitivity of the Mirostat sampling process
+        /// </summary>
         float MirostatTau { get; set; }
+
+
+        /// <summary>
+        /// Determines whether to apply penalty for generating newline characters ("\n") in the generated text.
+        /// </summary>
         bool PenalizeNL { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the penalty applied to token presence in the generated text
+        /// </summary>
         float PresencePenalty { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the number of tokens to repeat at the end of the generated text.
+        /// </summary>
         int RepeatLastTokensCount { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the penalty applied for repeating tokens in the generated text
+        /// </summary>
         float RepeatPenalty { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the temperature parameter for temperature-based sampling. Higher values make output more random, while lower values make it more deterministic
+        /// </summary>
         float Temperature { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the parameter (z) used in the TFS (Top Few Sampling) strategy.
+        /// </summary>
         float TfsZ { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the number of tokens to keep from the input text when generating output.
+        /// </summary>
         int TokensKeep { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the maximum number of tokens to consider during top-k sampling
+        /// </summary>
         int TopK { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the cumulative probability threshold for top-p sampling
+        /// </summary>
         float TopP { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the typicality penalty applied during language model inference.
+        /// </summary>
         float TypicalP { get; set; }
     }
 }

LLamaStack.Core/Config/ISessionConfig.cs

@@ -2,16 +2,54 @@
 
 namespace LLamaStack.Core.Config
 {
+    /// <summary>
+    /// Interface for Session configurations
+    /// </summary>
     public interface ISessionConfig
     {
+        /// <summary>
+        /// Gets or sets the model name to open the session on
+        /// </summary>
         string Model { get; set; }
+
+        /// <summary>
+        /// Gets or sets the type of the executor to use for inference.
+        /// </summary>
         ExecutorType ExecutorType { get; set; }
+
+        /// <summary>
+        /// Gets or sets the initial prompt to start the session with.
+        /// </summary>
         string Prompt { get; set; }
+
+        /// <summary>
+        /// Gets or sets the input prefix for Instruct executors.
+        /// </summary>
         string InputPrefix { get; set; }
+
+        /// <summary>
+        /// Gets or sets the input suffix for Instruct executors.
+        /// </summary>
         string InputSuffix { get; set; }
+
+        /// <summary>
+        /// Gets or sets one or more anti-prompt words as CSV. (Combined with AntiPrompts)
+        /// </summary>
         string AntiPrompt { get; set; }
+
+        /// <summary>
+        /// Gets or sets a list of anti-prompt words. (Combined with AntiPrompt)
+        /// </summary>
         public List<string> AntiPrompts { get; set; }
+
+        /// <summary>
+        /// Gets or sets a list of words to remove from the output as CSV. (Combined with OutputFilters)
+        /// </summary>
         string OutputFilter { get; set; }
+
+        /// <summary>
+        /// Gets or sets a list of words to remove from the output, (Combined with OutputFilter)
+        /// </summary>
         public List<string> OutputFilters { get; set; }
     }
 }

LLamaStack.Core/Config/InferenceConfig.cs

@@ -3,23 +3,91 @@
 
 namespace LLamaStack.Core.Config
 {
+    /// <summary>
+    /// Concrete implemtation of IInferenceConfig
+    /// </summary>
+    /// <seealso cref="LLamaStack.Core.Config.IInferenceConfig" />
     public class InferenceConfig : IInferenceConfig
     {
+        /// <summary>
+        /// Gets or sets the number of tokens to keep from the input text when generating output.
+        /// </summary>
         public int TokensKeep { get; set; } = 0;
+
+
+        /// <summary>
+        /// Gets or sets the maximum number of tokens to generate during inference, limiting the length of the generated text
+        /// </summary>
         public int MaxTokens { get; set; } = -1;
+
+        /// <summary>
+        /// Gets or sets the maximum number of tokens to consider during top-k sampling
+        /// </summary>
         public int TopK { get; set; } = 40;
+
+        /// <summary>
+        /// Gets or sets the cumulative probability threshold for top-p sampling
+        /// </summary>
         public float TopP { get; set; } = 0.95f;
+
+        /// <summary>
+        /// Gets or sets the parameter (z) used in the TFS (Top Few Sampling) strategy.
+        /// </summary>
         public float TfsZ { get; set; } = 1.0f;
+
+        /// <summary>
+        /// Gets or sets the typicality penalty applied during language model inference.
+        /// </summary>
         public float TypicalP { get; set; } = 1.0f;
+
+        /// <summary>
+        /// Gets or sets the temperature parameter for temperature-based sampling. Higher values make output more random, while lower values make it more deterministic
+        /// </summary>
         public float Temperature { get; set; } = 0.8f;
+
+        /// <summary>
+        /// Gets or sets the penalty applied for repeating tokens in the generated text
+        /// </summary>
         public float RepeatPenalty { get; set; } = 1.1f;
+
+        /// <summary>
+        /// Gets or sets the number of tokens to repeat at the end of the generated text.
+        /// </summary>
         public int RepeatLastTokensCount { get; set; } = 64;
+
+        /// <summary>
+        /// Gets or sets the penalty applied to token frequency, affecting token selection during language model inference.
+        /// </summary>
         public float FrequencyPenalty { get; set; } = .0f;
+
+        /// <summary>
+        /// Gets or sets the penalty applied to token presence in the generated text
+        /// </summary>
         public float PresencePenalty { get; set; } = .0f;
+
+        /// <summary>
+        /// Gets or sets the temperature or sensitivity of the Mirostat sampling process
+        /// </summary>
         public float MirostatTau { get; set; } = 5.0f;
+
+        /// <summary>
+        /// Gets or sets the mirostat eta used to adjust the strength of the Mirostat bias
+        /// </summary>
         public float MirostatEta { get; set; } = 0.1f;
+
+        /// <summary>
+        /// Determines whether to apply penalty for generating newline characters ("\n") in the generated text.
+        /// </summary>
         public bool PenalizeNL { get; set; } = true;
+
+        /// <summary>
+        /// Gets or sets the type of sampling strategy to use during language model inference (e.g., greedy, top-k, top-p)
+        /// </summary>
         public SamplerType SamplerType { get; set; } = SamplerType.Default;
+
+        /// <summary>
+        /// Gets or sets a list of <see cref="LLamaStack.Core.Models.LogitBiasModel" /> objects that provide bias information for specific tokens in the language model's vocabulary
+        /// </summary>
         public List<LogitBiasModel> LogitBias { get; set; } = new List<LogitBiasModel>();
     }
 }

LLamaStack.Core/Config/LLamaStackConfig.cs

@@ -2,12 +2,31 @@
 
 namespace LLamaStack.Core.Config
 {
+    /// <summary>
+    /// LLamaStack appsettings.json config element
+    /// </summary>
+    /// <seealso cref="LLamaStack.Core.Config.IConfigSection" />
     public class LLamaStackConfig : IConfigSection
     {
+        /// <summary>
+        /// Gets or sets the ModelLoad type
+        /// </summary>
         public ModelLoadType ModelLoadType { get; set; }
+
+        /// <summary>
+        /// Gets or sets the model state path.
+        /// </summary>
         public string ModelStatePath { get; set; }
+
+
+        /// <summary>
+        /// Gets or sets the models.
+        /// </summary>
         public List<ModelConfig> Models { get; set; }
 
+        /// <summary>
+        /// Perform any initialization, called directly after deserialization
+        /// </summary>
         public void Initialize()
         {
             if (string.IsNullOrEmpty(ModelStatePath))