From: Svjatoslav Agejenko <svjatoslav@svjatoslav.eu>
Date: Tue, 2 Dec 2025 19:52:26 +0000 (+0200)
Subject: Mark certain configuration parameters as optional in documentation.
X-Git-Url: http://www2.svjatoslav.eu/gitweb/?a=commitdiff_plain;h=5a54c387402c2c31a45292d3f3c7774edd24fc19;p=alyverkko-cli.git

Mark certain configuration parameters as optional in documentation.
---

diff --git a/doc/index.org b/doc/index.org
index 87ab3e7..0b47a99 100644
--- a/doc/index.org
+++ b/doc/index.org
@@ -359,22 +359,22 @@ Configuration file should be placed under current user home directory:
 
 **** Generation Parameters
 
-- =default_temperature=: Creativity control (0-3, higher = more
-  creative).
+- =default_temperature=: (Optional) Creativity control (0-3, higher =
+  more creative).
 
-- =default_top_p=: Nucleus sampling threshold (0.0-1., higher = more
-  diverse).
+- =default_top_p=: (Optional) Nucleus sampling threshold (0.0-1.,
+  higher = more diverse).
 
-- =default_top_k=: Restricts token selection to the K tokens with the
-  highest probabilities, regardless of their actual probability values
-  or the shape of the distribution.
+- =default_top_k=: (Optional) Restricts token selection to the K
+  tokens with the highest probabilities, regardless of their actual
+  probability values or the shape of the distribution.
 
-- =default_min_p=: Filters the vocabulary to include only tokens whose
-  probability is at least a certain fraction (Min P) of the
-  probability of the most likely token.
+- =default_min_p=: (Optional) Filters the vocabulary to include only
+  tokens whose probability is at least a certain fraction (Min P) of
+  the probability of the most likely token.
 
-- =default_repeat_penalty=: Penalty for repetition (>0.0, 1.0 = no
-  penalty)
+- =default_repeat_penalty=: (Optional) Penalty for repetition (>0.0,
+  1.0 = no penalty)
 
 **** Performance Tuning
 
@@ -395,18 +395,27 @@ Each model in the =models= list can have:
 - =alias=: Short model alias. Model with alias "default" would be used
   by default.
 
+- =temperature=: (Optional) (TODO: document)
+
+- =top_p=: (Optional) (TODO: document)
+
+- =min_p=: (Optional) (TODO: document)
+
+- =top_k=: (Optional) (TODO: document)
+
+- =repeat_penalty=: (Optional) (TODO: document)
+
 - =filesystem_path=: File name of the model as located within
   *models_directory*
 
 - =context_size_tokens=: Context size in tokens that model was trained
   on.
 
-- =end_of_text_marker=: Some models produce certain markers to
-  indicate end of their output. If specified here, Älyverkko CLI can
-  identify and remove them so that they don't leak into
+- =end_of_text_marker=: (Optional) Some models produce certain markers
+  to indicate end of their output. If specified here, Älyverkko CLI
+  can identify and remove them so that they don't leak into
   conversation. Default value is: *null*.
 
-- =temperature=, =top_p=, =repeat_penalty=: Model-specific overrides
 
 *** Configuration file example
 
@@ -419,10 +428,12 @@ file. Below is an example of how the configuration file might look:
   skills_directory: "/home/user/AI/skills"
   llama_cli_path: "/home/user/AI/llama.cpp/build/bin/llama-cli"
 
-  # Processing parameters
+  # Generation parameters
   default_temperature: 0.7
   default_top_p: 0.9
   default_repeat_penalty: 1.0
+
+  # Performance tuning
   thread_count: 6
   batch_thread_count: 10
 
@@ -431,20 +442,20 @@ file. Below is an example of how the configuration file might look:
     - alias: "default"
       filesystem_path: "model.gguf"
       context_size_tokens: 64000
-      end_of_text_marker: null
-      temperature: 0.8     # Optional model-specific parameter
-      top_p: 0.9           # Optional
-      repeat_penalty: 1.1  # Optional
+      temperature: 0.8
+      top_p: 0.9
+      repeat_penalty: 1.1
+
     - alias: "mistral"
       filesystem_path: "Mistral-Large-Instruct-2407.Q8_0.gguf"
       context_size_tokens: 32768
-      end_of_text_marker: null
 #+end_src
 
 *** Parameter Precedence Hierarchy
 
-For *temperature*, *top_p*, and *repeat_penalty* parameters, values
-are determined using this priority order (highest to lowest):
+For *temperature*, *top_p*, *top_k*, *min_p* and *repeat_penalty*
+parameters, values are determined using this priority order (highest
+to lowest):
 
 1. *Skill-specific value* (from skill's YAML file)
 2. *Model-specific value* (from model configuration)
diff --git a/src/main/java/eu/svjatoslav/alyverkko_cli/commands/task_processor/TaskProcess.java b/src/main/java/eu/svjatoslav/alyverkko_cli/commands/task_processor/TaskProcess.java
index a228819..e266763 100644
--- a/src/main/java/eu/svjatoslav/alyverkko_cli/commands/task_processor/TaskProcess.java
+++ b/src/main/java/eu/svjatoslav/alyverkko_cli/commands/task_processor/TaskProcess.java
@@ -207,9 +207,6 @@ public class TaskProcess {
 
     }
 
-
-
-
     /**
      * Spawns a new Thread to handle the error stream from llama.cpp,
      * printing lines that contain metadata or errors to the console.