Webui stuff

Author: Silver267

common/common.h

@@ -136,12 +136,12 @@ struct common_params_sampling {
     float   min_p              = 0.05f; // 0.0 = disabled
     float   xtc_probability    = 0.00f; // 0.0 = disabled
     float   xtc_threshold      = 0.10f; // > 0.5 disables XTC
+    float   smoothing_factor   = 0.0f;  // controls the quadratic adjustment in smooth / quadratic sampling (0.0 = disabled)
+    float   smoothing_curve    = 1.0f;  // controls the cubic transformation curve for smoothing / quadratic sampling.
     float   typ_p              = 1.00f; // typical_p, 1.0 = disabled
     float   temp               = 0.80f; // <= 0.0 to sample greedily, 0.0 to not output probabilities
     float   dynatemp_range     = 0.00f; // 0.0 = disabled
     float   dynatemp_exponent  = 1.00f; // controls how entropy maps to temperature in dynamic temperature sampler
-    float   smoothing_factor   = 0.0f;  // controls the quadratic adjustment in smooth / quadratic sampling (0.0 = disabled)
-    float   smoothing_curve    = 1.0f;  // controls the quadratic adjustment in smooth / quadratic sampling
     int32_t penalty_last_n     = 64;    // last n tokens to penalize (0 = disable penalty, -1 = context size)
     float   penalty_repeat     = 1.00f; // 1.0 = disabled
     float   penalty_freq       = 0.00f; // 0.0 = disabled

tools/main/README.md

@@ -309,7 +309,7 @@ By performing a non-linear transformation on token logits, we can effectively av
 
 `smoothing curve` is a second hyperparameter that adds a cubic transformation on top of the original quadratic one, and can "help make lower `smoothing factor` values work if the curve is set higher. A smoothing curve value of `1.0` is equivalant of using just quadratic transformation.
 
-This sampler is not mutually exclusive with Temperature, they can be used together. 
+This sampler is not mutually exclusive with Temperature, they can be used together.
 
 ### Top-nσ Sampling
 

tools/server/README.md

@@ -105,7 +105,7 @@ The project is under active development, and we are [looking for feedback and co
 
 | Argument | Explanation |
 | -------- | ----------- |
-| `--samplers SAMPLERS` | samplers that will be used for generation in the order, separated by ';'<br/>(default: penalties;dry;top_n_sigma;top_k;typ_p;top_p;min_p;xtc;temperature) |
+| `--samplers SAMPLERS` | samplers that will be used for generation in the order, separated by ';'<br/>(default: penalties;dry;top_n_sigma;top_k;typ_p;top_p;min_p;xtc;smoothing;temperature) |
 | `-s, --seed SEED` | RNG seed (default: -1, use random seed for -1) |
 | `--sampling-seq, --sampler-seq SEQUENCE` | simplified sequence for samplers that will be used (default: edskypmxt) |
 | `--ignore-eos` | ignore end of stream token and continue generating (implies --logit-bias EOS-inf) |
@@ -483,7 +483,7 @@ These words will not be included in the completion, so make sure to add them to
 
 `return_tokens`: Return the raw generated token ids in the `tokens` field. Otherwise `tokens` remains empty. Default: `false`
 
-`samplers`: The order the samplers should be applied in. An array of strings representing sampler type names. If a sampler is not set, it will not be used. If a sampler is specified more than once, it will be applied multiple times. Default: `["dry", "top_k", "typ_p", "top_p", "min_p", "xtc", "temperature"]` - these are all the available values.
+`samplers`: The order the samplers should be applied in. An array of strings representing sampler type names. If a sampler is not set, it will not be used. If a sampler is specified more than once, it will be applied multiple times. Default: `["dry", "top_k", "typ_p", "top_p", "min_p", "xtc", "smoothing", "temperature"]` - these are all the available values.
 
 `timings_per_token`: Include prompt processing and text generation speed information in each response.  Default: `false`
 
@@ -732,6 +732,8 @@ This endpoint is public (no API key check). By default, it is read-only. To make
       "min_p": 0.05000000074505806,
       "xtc_probability": 0.0,
       "xtc_threshold": 0.10000000149011612,
+      "smoothing_factor": 0.0,
+      "smoothing_curve": 1.0,
       "typical_p": 1.0,
       "repeat_last_n": 64,
       "repeat_penalty": 1.0,
@@ -766,6 +768,7 @@ This endpoint is public (no API key check). By default, it is read-only. To make
         "top_p",
         "min_p",
         "xtc",
+        "smoothing",
         "temperature"
       ],
       "speculative.n_max": 16,
@@ -878,6 +881,8 @@ Example:
       "min_p": 0.05000000074505806,
       "xtc_probability": 0.0,
       "xtc_threshold": 0.10000000149011612,
+      "smoothing_factor": 0.0,
+      "smoothing_curve": 1.0,
       "typical_p": 1.0,
       "repeat_last_n": 64,
       "repeat_penalty": 1.0,
@@ -912,6 +917,7 @@ Example:
         "top_p",
         "min_p",
         "xtc",
+        "smoothing",
         "temperature"
       ],
       "speculative.n_max": 16,

tools/server/public/index.html.gz

@@ -48,6 +48,8 @@
       min_p: 0.05, // 0 = disabled; recommended for non-english: ~ 0.4
       xtc_probability: 0.0, // 0 = disabled;
       xtc_threshold: 0.1, // > 0.5 disables XTC;
+      smoothing_factor: 0.0, // 0 = disabled;
+      smoothing_curve: 1.0, // > 0.5 = disabled
       typical_p: 1.0, // 1.0 = disabled
       presence_penalty: 0.0, // 0.0 = disabled
       frequency_penalty: 0.0, // 0.0 = disabled
@@ -841,6 +843,8 @@
           ${FloatField({ label: "Typical-P", title: "Activates local typical sampling, a method used to limit the prediction of tokens that are atypical in the current context. The parameter p controls the strength of this limitation. A value of 1.0 means that this function is deactivated.", max: 1.0, min: 0.0, name: "typical_p", step: 0.01, value: params.value.typical_p })}
           ${FloatField({ label: "XTC probability", title: "Sets the chance for token removal (checked once on sampler start)", max: 1.0, min: 0.0, name: "xtc_probability", step: 0.01, value: params.value.xtc_probability })}
           ${FloatField({ label: "XTC threshold", title: "Sets a minimum probability threshold for tokens to be removed", max: 0.5, min: 0.0, name: "xtc_threshold", step: 0.01, value: params.value.xtc_threshold })}
+          ${FloatField({ label: "Smoothing Factor", max: 10.0, min: 0.0, name: "smoothing_factor", step: 0.01, value: params.value.smoothing_factor })}
+          ${FloatField({ label: "Smoothing Curve", max: 10.0, min: 0.0, name: "smoothing_curve", step: 0.01, value: params.value.smoothing_curve })}
           ${FloatField({ label: "DRY Penalty Multiplier", title: "Set the DRY repetition penalty multiplier. Default is 0.0, which disables DRY.", max: 5.0, min: 0.0, name: "dry_multiplier", step: 0.01, value: params.value.dry_multiplier })}
           ${FloatField({ label: "DRY Base", title: "Set the DRY repetition penalty base value. Default is 1.75", max: 3.0, min: 1.0, name: "dry_base", step: 0.01, value: params.value.dry_base })}
           ${IntField({ label: "DRY Allowed Length", title: "Tokens that extend repetition beyond this receive exponentially increasing penalty. Default is 2", max: 10, min: 1, step: 1, name: "dry_allowed_length", value: params.value.dry_allowed_length })}
@@ -1143,6 +1147,8 @@ <h2>llama.cpp</h2>
     min_p: { snapValue: 0.05, snapRangeMultiplier: 2 },
     xtc_probability: { snapValue: 0.0, snapRangeMultiplier: 4 },
     xtc_threshold: { snapValue: 0.5, snapRangeMultiplier: 4 },
+    smoothing_factor: { snapValue: 0.0, snapRangeMultiplier: 4 },
+    smoothing_curve: { snapValue: 0.0, snapRangeMultiplier: 4 },
     top_p: { snapValue: 1.0, snapRangeMultiplier: 4 },
     typical_p: { snapValue: 1.0, snapRangeMultiplier: 4 },
     repeat_penalty: { snapValue: 1.0, snapRangeMultiplier: 4 },

tools/server/public_legacy/index-new.html

@@ -312,6 +312,8 @@
       min_p: 0.05, // 0 = disabled
       xtc_probability: 0.0, // 0 = disabled;
       xtc_threshold: 0.1, // > 0.5 disables XTC;
+      smoothing_factor: 0.0, // 0 = disabled;
+      smoothing_curve: 1.0, // > 0.5 = disabled
       typical_p: 1.0, // 1.0 = disabled
       presence_penalty: 0.0, // 0.0 = disabled
       frequency_penalty: 0.0, // 0.0 = disabled
@@ -1021,6 +1023,8 @@
               ${IntField({ label: "DRY Penalty Last N", max: 2048, min: -1, step: 16, name: "dry_penalty_last_n", value: params.value.dry_penalty_last_n })}
               ${FloatField({ label: "XTC probability", max: 1.0, min: 0.0, name: "xtc_probability", step: 0.01, value: params.value.xtc_probability })}
               ${FloatField({ label: "XTC threshold", max: 0.5, min: 0.0, name: "xtc_threshold", step: 0.01, value: params.value.xtc_threshold })}
+              ${FloatField({ label: "Smoothing Factor", max: 10.0, min: 0.0, name: "smoothing_factor", step: 0.01, value: params.value.smoothing_factor })}
+              ${FloatField({ label: "Smoothing Curve", max: 10.0, min: 0.0, name: "smoothing_curve", step: 0.01, value: params.value.smoothing_curve })}
             </fieldset>
             <hr />
             <fieldset class="three">

tools/server/public_legacy/index.html

@@ -144,14 +144,14 @@ struct slot_params {
             {"temperature",               sampling.temp},
             {"dynatemp_range",            sampling.dynatemp_range},
             {"dynatemp_exponent",         sampling.dynatemp_exponent},
-            {"smoothing_factor",          sampling.smoothing_factor},
-            {"smoothing_curve",           sampling.smoothing_curve},
             {"top_k",                     sampling.top_k},
             {"top_p",                     sampling.top_p},
             {"min_p",                     sampling.min_p},
             {"top_n_sigma",               sampling.top_n_sigma},
             {"xtc_probability",           sampling.xtc_probability},
             {"xtc_threshold",             sampling.xtc_threshold},
+            {"smoothing_factor",          sampling.smoothing_factor},
+            {"smoothing_curve",           sampling.smoothing_curve},
             {"typical_p",                 sampling.typ_p},
             {"repeat_last_n",             sampling.penalty_last_n},
             {"repeat_penalty",            sampling.penalty_repeat},
@@ -255,12 +255,12 @@ struct server_task {
         params.sampling.top_n_sigma        = json_value(data, "top_n_sigma",        defaults.sampling.top_n_sigma);
         params.sampling.xtc_probability    = json_value(data, "xtc_probability",    defaults.sampling.xtc_probability);
         params.sampling.xtc_threshold      = json_value(data, "xtc_threshold",      defaults.sampling.xtc_threshold);
+        params.sampling.smoothing_factor   = json_value(data, "smoothing_factor",   defaults.sampling.smoothing_factor);
+        params.sampling.smoothing_curve    = json_value(data, "smoothing_curve",    defaults.sampling.smoothing_curve);
         params.sampling.typ_p              = json_value(data, "typical_p",          defaults.sampling.typ_p);
         params.sampling.temp               = json_value(data, "temperature",        defaults.sampling.temp);
         params.sampling.dynatemp_range     = json_value(data, "dynatemp_range",     defaults.sampling.dynatemp_range);
         params.sampling.dynatemp_exponent  = json_value(data, "dynatemp_exponent",  defaults.sampling.dynatemp_exponent);
-        params.sampling.smoothing_factor   = json_value(data, "smoothing_factor",   defaults.sampling.smoothing_factor);
-        params.sampling.smoothing_curve    = json_value(data, "smoothing_curve",    defaults.sampling.smoothing_curve);
         params.sampling.penalty_last_n     = json_value(data, "repeat_last_n",      defaults.sampling.penalty_last_n);
         params.sampling.penalty_repeat     = json_value(data, "repeat_penalty",     defaults.sampling.penalty_repeat);
         params.sampling.penalty_freq       = json_value(data, "frequency_penalty",  defaults.sampling.penalty_freq);

tools/server/server.cpp

@@ -17,7 +17,7 @@ export const CONFIG_DEFAULT = {
   showThoughtInProgress: false,
   excludeThoughtOnReq: true,
   // make sure these default values are in sync with `common.h`
-  samplers: 'edkypmxt',
+  samplers: 'edkypmxqt',
   temperature: 0.8,
   dynatemp_range: 0.0,
   dynatemp_exponent: 1.0,
@@ -26,6 +26,8 @@ export const CONFIG_DEFAULT = {
   min_p: 0.05,
   xtc_probability: 0.0,
   xtc_threshold: 0.1,
+  smoothing_factor: 0.0,
+  smoothing_curve: 1.0,
   typical_p: 1.0,
   repeat_last_n: 64,
   repeat_penalty: 1.0,
@@ -44,7 +46,7 @@ export const CONFIG_INFO: Record<string, string> = {
   apiKey: 'Set the API Key if you are using --api-key option for the server.',
   systemMessage: 'The starting message that defines how model should behave.',
   samplers:
-    'The order at which samplers are applied, in simplified way. Default is "dkypmxt": dry->top_k->typ_p->top_p->min_p->xtc->temperature',
+    'The order at which samplers are applied, in simplified way. Default is "dkypmxqt": dry->top_k->typ_p->top_p->min_p->xtc->smoothing->temperature',
   temperature:
     'Controls the randomness of the generated text by affecting the probability distribution of the output tokens. Higher = more random, lower = more focused.',
   dynatemp_range:
@@ -60,6 +62,10 @@ export const CONFIG_INFO: Record<string, string> = {
     'XTC sampler cuts out top tokens; this parameter controls the chance of cutting tokens at all. 0 disables XTC.',
   xtc_threshold:
     'XTC sampler cuts out top tokens; this parameter controls the token probability that is required to cut that token.',
+  smoothing_factor:
+    'Controls the quadratic adjustment in smooth / quadratic sampler',
+  smoothing_curve:
+    'Controls the cubic transformation curve for smoothing / quadratic sampler.',
   typical_p:
     'Sorts and limits tokens based on the difference between log-probability and entropy.',
   repeat_last_n: 'Last n tokens to consider for penalizing repetition',

tools/server/webui/src/Config.ts

@@ -29,6 +29,8 @@ const SAMPLER_KEYS: SettKey[] = [
   'typical_p',
   'xtc_probability',
   'xtc_threshold',
+  'smoothing_factor',
+  'smoothing_curve',
 ];
 const PENALTY_KEYS: SettKey[] = [
   'repeat_last_n',

tools/server/webui/src/components/SettingDialog.tsx

@@ -219,6 +219,8 @@ export const AppContextProvider = ({
         typical_p: config.typical_p,
         xtc_probability: config.xtc_probability,
         xtc_threshold: config.xtc_threshold,
+        smoothing_factor: config.smoothing_factor,
+        smoothing_curve: config.smoothing_curve,
         repeat_last_n: config.repeat_last_n,
         repeat_penalty: config.repeat_penalty,
         presence_penalty: config.presence_penalty,