examples : add --print-confidence option to cli (#3150)

* examples : add --print-confidence option to cli

This commit adds a new command-line option `--print-confidence` to the
whisper-cli. When enabled, this option prints the confidence level of each
token in the transcribed text using ANSI formatting codes.

The confidence levels are represented using different styles:
```console
main: confidence: highlighted (low confidence), underlined (medium), dim (high confidence)
```

Refs: https://github.com/ggml-org/whisper.cpp/issues/3135

diff --git a/examples/cli/cli.cpp b/examples/cli/cli.cpp
index 7d5d0ffea3590a989a05bf19d77e355e0b1d2a2b..2a0f940d5d739cbef08420e77cb2312cbbd450f3 100644 (file)
--- a/examples/cli/cli.cpp
+++ b/examples/cli/cli.cpp
@@ -70,6 +70,7 @@ struct whisper_params {
     bool no_prints       = false;
     bool print_special   = false;
     bool print_colors    = false;
+    bool print_confidence= false;
     bool print_progress  = false;
     bool no_timestamps   = false;
     bool log_score       = false;
@@ -179,6 +180,7 @@ static bool whisper_params_parse(int argc, char ** argv, whisper_params & params
         else if (arg == "-np"   || arg == "--no-prints")       { params.no_prints       = true; }
         else if (arg == "-ps"   || arg == "--print-special")   { params.print_special   = true; }
         else if (arg == "-pc"   || arg == "--print-colors")    { params.print_colors    = true; }
+        else if (                  arg == "--print-confidence"){ params.print_confidence= true; }
         else if (arg == "-pp"   || arg == "--print-progress")  { params.print_progress  = true; }
         else if (arg == "-nt"   || arg == "--no-timestamps")   { params.no_timestamps   = true; }
         else if (arg == "-l"    || arg == "--language")        { params.language        = whisper_param_turn_lowercase(ARGV_NEXT); }
@@ -257,6 +259,7 @@ static void whisper_print_usage(int /*argc*/, char ** argv, const whisper_params
     fprintf(stderr, "  -np,       --no-prints         [%-7s] do not print anything other than the results\n",   params.no_prints ? "true" : "false");
     fprintf(stderr, "  -ps,       --print-special     [%-7s] print special tokens\n",                           params.print_special ? "true" : "false");
     fprintf(stderr, "  -pc,       --print-colors      [%-7s] print colors\n",                                   params.print_colors ? "true" : "false");
+    fprintf(stderr, "             --print-confidence  [%-7s] print confidence\n",                               params.print_confidence ? "true" : "false");
     fprintf(stderr, "  -pp,       --print-progress    [%-7s] print progress\n",                                 params.print_progress ? "true" : "false");
     fprintf(stderr, "  -nt,       --no-timestamps     [%-7s] do not print timestamps\n",                        params.no_timestamps ? "true" : "false");
     fprintf(stderr, "  -l LANG,   --language LANG     [%-7s] spoken language ('auto' for auto-detect)\n",       params.language.c_str());
@@ -386,6 +389,26 @@ static void whisper_print_segment_callback(struct whisper_context * ctx, struct
 
                 printf("%s%s%s%s", speaker.c_str(), k_colors[col].c_str(), text, "\033[0m");
             }
+        } else if (params.print_confidence) {
+            for (int j = 0; j < whisper_full_n_tokens(ctx, i); ++j) {
+                if (params.print_special == false) {
+                    const whisper_token id = whisper_full_get_token_id(ctx, i, j);
+                    if (id >= whisper_token_eot(ctx)) {
+                        continue;
+                    }
+                }
+
+                const char * text = whisper_full_get_token_text(ctx, i, j);
+                const float  p    = whisper_full_get_token_p   (ctx, i, j);
+
+                int style_idx = 2;     // High confidence - dim
+                if (p < 0.33) {
+                    style_idx = 0;     // Low confidence - inverse (highlighted)
+                } else if (p < 0.66) {
+                    style_idx = 1;     // Medium confidence - underlined
+                }
+                printf("%s%s%s%s", speaker.c_str(), k_styles[style_idx].c_str(), text, "\033[0m");
+            }
         } else {
             const char * text = whisper_full_get_segment_text(ctx, i);
 
@@ -1115,6 +1138,8 @@ int main(int argc, char ** argv) {
 
             if (params.print_colors) {
                 fprintf(stderr, "%s: color scheme: red (low confidence), yellow (medium), green (high confidence)\n", __func__);
+            } else if (params.print_confidence) {
+                fprintf(stderr, "%s: confidence: highlighted (low confidence), underlined (medium), dim (high confidence)\n", __func__);
             }
             fprintf(stderr, "\n");
         }

diff --git a/examples/common.h b/examples/common.h
index 1aa763817ee93261f9d7b48381e2693e4410bd7e..8f99df7dd326a0f6587a9b64628475098b75a656 100644 (file)
--- a/examples/common.h
+++ b/examples/common.h
@@ -294,6 +294,26 @@ const std::vector<std::string> k_colors = {
     set_xterm256_foreground( 78, 178, 101),
 };
 
+// ANSI formatting codes
+static std::string set_inverse() {
+    return "\033[7m";
+}
+
+static std::string set_underline() {
+    return "\033[4m";
+}
+
+static std::string set_dim() {
+    return "\033[2m";
+}
+
+// Style scheme for different confidence levels
+const std::vector<std::string> k_styles = {
+    set_inverse(),   // Low confidence - inverse (highlighted)
+    set_underline(), // Medium confidence - underlined
+    set_dim(),       // High confidence - dim
+};
+
 //
 // Other utils
 //