whispercpp 1.3.2 → 1.3.4

data/ext/sources/examples/talk-llama/talk-llama.cpp

@@ -66,7 +66,7 @@ struct whisper_params {
     float top_p = 0.80f;
     float min_p = 0.01f;
     float temp  = 0.30f;
-    
+
     float vad_thold  = 0.6f;
     float freq_thold = 100.0f;
 
@@ -76,7 +76,7 @@ struct whisper_params {
     bool no_timestamps  = true;
     bool verbose_prompt = false;
     bool use_gpu        = true;
-    bool flash_attn     = false;
+    bool flash_attn     = true;
 
     std::string person      = "Georgi";
     std::string bot_name    = "LLaMA";
@@ -122,6 +122,7 @@ static bool whisper_params_parse(int argc, char ** argv, whisper_params & params
         else if (arg == "-vp"  || arg == "--verbose-prompt") { params.verbose_prompt = true; }
         else if (arg == "-ng"  || arg == "--no-gpu")         { params.use_gpu        = false; }
         else if (arg == "-fa"  || arg == "--flash-attn")     { params.flash_attn     = true; }
+        else if (arg == "-nfa" || arg == "--no-flash-attn")  { params.flash_attn     = false; }
         else if (arg == "-p"   || arg == "--person")         { params.person         = argv[++i]; }
         else if (arg == "-bn"   || arg == "--bot-name")      { params.bot_name       = argv[++i]; }
         else if (arg == "--session")                         { params.path_session   = argv[++i]; }
@@ -175,7 +176,8 @@ void whisper_print_usage(int /*argc*/, char ** argv, const whisper_params & para
     fprintf(stderr, "  -pe,      --print-energy   [%-7s] print sound energy (for debugging)\n",          params.print_energy ? "true" : "false");
     fprintf(stderr, "  -vp,      --verbose-prompt [%-7s] print prompt at start\n",                       params.verbose_prompt ? "true" : "false");
     fprintf(stderr, "  -ng,      --no-gpu         [%-7s] disable GPU\n",                                 params.use_gpu ? "false" : "true");
-    fprintf(stderr, "  -fa,      --flash-attn     [%-7s] flash attention\n",                             params.flash_attn ? "true" : "false");
+    fprintf(stderr, "  -fa,      --flash-attn     [%-7s] enable flash attention\n",                      params.flash_attn ? "true" : "false");
+    fprintf(stderr, "  -nfa,     --no-flash-attn  [%-7s] disable flash attention\n",                     params.flash_attn ? "false" : "true");
     fprintf(stderr, "  -p NAME,  --person NAME    [%-7s] person name (for prompt selection)\n",          params.person.c_str());
     fprintf(stderr, "  -bn NAME, --bot-name NAME  [%-7s] bot name (to display)\n",                       params.bot_name.c_str());
     fprintf(stderr, "  -w TEXT,  --wake-command T [%-7s] wake-up command to listen for\n",               params.wake_cmd.c_str());
@@ -291,6 +293,8 @@ The transcript only includes text, it does not include markup like HTML and Mark
 {0}{4})";
 
 int main(int argc, char ** argv) {
+    ggml_backend_load_all();
+
     whisper_params params;
 
     if (whisper_params_parse(argc, argv, params) == false) {
@@ -338,9 +342,10 @@ int main(int argc, char ** argv) {
     llama_context_params lcparams = llama_context_default_params();
 
     // tune these to your liking
-    lcparams.n_ctx      = 2048;
-    lcparams.n_threads  = params.n_threads;
-    lcparams.flash_attn = params.flash_attn;
+    lcparams.n_ctx     = 2048;
+    lcparams.n_threads = params.n_threads;
+
+    lcparams.flash_attn_type = params.flash_attn ? LLAMA_FLASH_ATTN_TYPE_AUTO : LLAMA_FLASH_ATTN_TYPE_DISABLED;
 
     struct llama_context * ctx_llama = llama_init_from_model(model_llama, lcparams);
 

data/ext/sources/examples/talk-llama/unicode.cpp

@@ -204,12 +204,17 @@ static inline std::wstring unicode_wstring_from_utf8(const std::string & s) {
     // disable C++17 deprecation warning for std::codecvt_utf8
 #    pragma clang diagnostic push
 #    pragma clang diagnostic ignored "-Wdeprecated-declarations"
+#elif defined(__GNUC__)
+#    pragma GCC diagnostic push
+#    pragma GCC diagnostic ignored "-Wdeprecated-declarations"
 #endif
 
     std::wstring_convert<std::codecvt_utf8<wchar_t>> conv;
 
 #if defined(__clang__)
 #    pragma clang diagnostic pop
+#elif defined(__GNUC__)
+#    pragma GCC diagnostic pop
 #endif
 
     return conv.from_bytes(s);
@@ -552,6 +557,178 @@ static std::vector<size_t> unicode_regex_split_stl(const std::string & text, con
     return bpe_offsets;
 }
 
+// K2 system regex patterns (from tokenization_kimi.py):
+// [\p{Han}]+|[^\r\n\p{L}\p{N}]?[\p{Lu}\p{Lt}\p{Lm}\p{Lo}\p{M}&&[^\p{Han}]]*[\p{Ll}\p{Lm}\p{Lo}\p{M}&&[^\p{Han}]]+(?i:'s|'t|'re|'ve|'m|'ll|'d)?|[^\r\n\p{L}\p{N}]?[\p{Lu}\p{Lt}\p{Lm}\p{Lo}\p{M}&&[^\p{Han}]]+[\p{Ll}\p{Lm}\p{Lo}\p{M}&&[^\p{Han}]]*(?i:'s|'t|'re|'ve|'m|'ll|'d)?|\p{N}{1,3}| ?[^\s\p{L}\p{N}]+[\r\n]*|\s*[\r\n]+|\s+(?!\S)|\s+
+static std::vector<size_t> unicode_regex_split_custom_kimi_k2(const std::string & text, const std::vector<size_t> & offsets) {
+    std::vector<size_t> bpe_offsets;
+    bpe_offsets.reserve(offsets.size());
+
+    const auto cpts = unicode_cpts_from_utf8(text);
+
+    size_t start = 0;
+    for (auto offset : offsets) {
+        const size_t offset_ini = start;
+        const size_t offset_end = start + offset;
+        assert(offset_end <= cpts.size());
+        start = offset_end;
+
+        static const uint32_t OUT_OF_RANGE = 0xFFFFFFFF;
+        auto _get_cpt = [&] (const size_t pos) -> uint32_t {
+            return (offset_ini <= pos && pos < offset_end) ? cpts[pos] : OUT_OF_RANGE;
+        };
+
+        auto _get_flags = [&] (const size_t pos) -> unicode_cpt_flags {
+            return (offset_ini <= pos && pos < offset_end) ? unicode_cpt_flags_from_cpt(cpts[pos]) : unicode_cpt_flags{};
+        };
+
+        size_t _prev_end = offset_ini;
+        auto _add_token = [&] (const size_t end) -> size_t {
+            assert(_prev_end <= end && end <= offset_end);
+            size_t len = end - _prev_end;
+            if (len > 0) {
+                bpe_offsets.push_back(len);
+            }
+            _prev_end = end;
+            return len;
+        };
+
+        for (size_t pos = offset_ini; pos < offset_end; /*pos++*/ ) {
+            const uint32_t cpt = _get_cpt(pos);
+            const auto flags = _get_flags(pos);
+
+            // Pattern 1: [\p{Han}]+ (Chinese characters)
+            if (unicode_cpt_is_han(cpt)) {
+                while (unicode_cpt_is_han(_get_cpt(pos))) {
+                    pos++;
+                }
+                _add_token(pos);
+                continue;
+            }
+
+            // Pattern 2 & 3: Letter words excluding Han characters with optional contractions
+            // [^\r\n\p{L}\p{N}]?[\p{Lu}\p{Lt}\p{Lm}\p{Lo}\p{M}&&[^\p{Han}]]*[\p{Ll}\p{Lm}\p{Lo}\p{M}&&[^\p{Han}]]+(?:'s|'t|'re|'ve|'m|'ll|'d)?
+            // [^\r\n\p{L}\p{N}]?[\p{Lu}\p{Lt}\p{Lm}\p{Lo}\p{M}&&[^\p{Han}]]+[\p{Ll}\p{Lm}\p{Lo}\p{M}&&[^\p{Han}]]*(?:'s|'t|'re|'ve|'m|'ll|'d)?
+            // Check if current char is a letter OR if current char could be a leading char and next char is a letter
+            bool is_letter_pattern = (flags.is_letter && !unicode_cpt_is_han(cpt)) ||
+                                     (!(cpt == '\r' || cpt == '\n' || flags.is_letter || flags.is_number) &&
+                                      _get_flags(pos + 1).is_letter && !unicode_cpt_is_han(_get_cpt(pos + 1)));
+
+            if (is_letter_pattern) {
+                // Handle optional leading non-letter/non-number character
+                bool has_leading_char = false;
+                if (!(cpt == '\r' || cpt == '\n' || flags.is_letter || flags.is_number)) {
+                    has_leading_char = true;
+                    pos++;
+                }
+
+                // Match letter sequence (excluding Han characters)
+                bool has_letters = false;
+                while (_get_flags(pos).is_letter && !unicode_cpt_is_han(_get_cpt(pos))) {
+                    has_letters = true;
+                    pos++;
+                }
+
+                // Only proceed if we found letters (after potentially skipping leading char)
+                if (has_letters || (!has_leading_char && _get_flags(pos).is_letter && !unicode_cpt_is_han(_get_cpt(pos)))) {
+                    if (!has_letters) pos++; // consume the first letter if we didn't already
+
+                    // Continue consuming letters
+                    while (_get_flags(pos).is_letter && !unicode_cpt_is_han(_get_cpt(pos))) {
+                        pos++;
+                    }
+
+                    // Check for optional contractions (?:'s|'t|'re|'ve|'m|'ll|'d)
+                    if (_get_cpt(pos) == '\'' && pos + 1 < offset_end) {
+                        uint32_t cpt_next = unicode_tolower(_get_cpt(pos + 1));
+                        if (cpt_next == 's' || cpt_next == 't' || cpt_next == 'm' || cpt_next == 'd') {
+                            pos += 2;
+                        } else if (pos + 2 < offset_end) {
+                            uint32_t cpt_next_next = unicode_tolower(_get_cpt(pos + 2));
+                            if ((cpt_next == 'r' && cpt_next_next == 'e') ||
+                                (cpt_next == 'v' && cpt_next_next == 'e') ||
+                                (cpt_next == 'l' && cpt_next_next == 'l')) {
+                                pos += 3;
+                            }
+                        }
+                    }
+
+                    _add_token(pos);
+                    continue;
+                } else if (has_leading_char) {
+                    // We consumed a leading char but found no letters, backtrack
+                    pos--;
+                }
+            }
+
+            // Pattern 4: \p{N}{1,3} (numbers 1-3 digits)
+            if (flags.is_number) {
+                size_t ini = pos;
+                while (_get_flags(pos).is_number) {
+                    if (++pos - ini >= 3) {
+                        _add_token(pos);
+                        ini = pos;
+                    }
+                }
+                _add_token(pos);
+                continue;
+            }
+
+            // Pattern 5:  ?[^\s\p{L}\p{N}]+[\r\n]* (optional space + non-word chars + optional newlines)
+            auto flags2 = (cpt == ' ' ? _get_flags(pos + 1) : flags);
+            if (!(flags2.is_whitespace || flags2.is_letter || flags2.is_number) && flags2.as_uint()) {
+                pos += (cpt == ' ');
+                while (!(flags2.is_whitespace || flags2.is_letter || flags2.is_number) && flags2.as_uint()) {
+                    flags2 = _get_flags(++pos);
+                }
+                // Match optional [\r\n]*
+                uint32_t cpt2 = _get_cpt(pos);
+                while (cpt2 == '\r' || cpt2 == '\n') {
+                    cpt2 = _get_cpt(++pos);
+                }
+                _add_token(pos);
+                continue;
+            }
+
+            // Count whitespace characters
+            size_t num_whitespaces = 0;
+            size_t last_end_r_or_n = 0;
+            while (_get_flags(pos + num_whitespaces).is_whitespace) {
+                uint32_t cpt2 = _get_cpt(pos + num_whitespaces);
+                if (cpt2 == '\r' || cpt2 == '\n') {
+                    last_end_r_or_n = pos + num_whitespaces + 1;
+                }
+                num_whitespaces++;
+            }
+
+            // Pattern 6: \s*[\r\n]+ (whitespace with newlines)
+            if (last_end_r_or_n > 0) {
+                pos = last_end_r_or_n;
+                _add_token(pos);
+                continue;
+            }
+
+            // Pattern 7: \s+(?!\S) (trailing whitespace)
+            if (num_whitespaces > 1 && _get_cpt(pos + num_whitespaces) != OUT_OF_RANGE) {
+                pos += num_whitespaces - 1;
+                _add_token(pos);
+                continue;
+            }
+
+            // Pattern 8: \s+ (general whitespace)
+            if (num_whitespaces > 0) {
+                pos += num_whitespaces;
+                _add_token(pos);
+                continue;
+            }
+
+            // No matches - consume single character
+            _add_token(++pos);
+        }
+    }
+
+    return bpe_offsets;
+}
+
 static std::vector<size_t> unicode_regex_split_custom(const std::string & text, const std::string & regex_expr, const std::vector<size_t> & offsets) {
     std::vector<size_t> bpe_offsets;
 
@@ -562,6 +739,9 @@ static std::vector<size_t> unicode_regex_split_custom(const std::string & text,
             regex_expr == "(?:'[sS]|'[tT]|'[rR][eE]|'[vV][eE]|'[mM]|'[lL][lL]|'[dD])|[^\\r\\n\\p{L}\\p{N}]?\\p{L}+|\\p{N}{1,3}| ?[^\\s\\p{L}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+") {
 
         bpe_offsets = unicode_regex_split_custom_llama3(text, offsets);
+    } else if (regex_expr == "\\p{Han}+") {
+        // K2's first pattern - handle all K2 patterns together
+        bpe_offsets = unicode_regex_split_custom_kimi_k2(text, offsets);
     }
 
     return bpe_offsets;
@@ -667,6 +847,38 @@ uint32_t unicode_tolower(uint32_t cpt) {
     return cpt;  // Return the original code point if no lowercase mapping is found
 }
 
+bool unicode_cpt_is_han(uint32_t cpt) {
+    // Han character ranges (Chinese/CJK characters)
+    // CJK Unified Ideographs (most common)
+    if (cpt >= 0x4E00 && cpt <= 0x9FFF) return true;
+
+    // CJK Extension A
+    if (cpt >= 0x3400 && cpt <= 0x4DBF) return true;
+
+    // CJK Extension B
+    if (cpt >= 0x20000 && cpt <= 0x2A6DF) return true;
+
+    // CJK Extension C
+    if (cpt >= 0x2A700 && cpt <= 0x2B73F) return true;
+
+    // CJK Extension D
+    if (cpt >= 0x2B740 && cpt <= 0x2B81F) return true;
+
+    // CJK Extension E
+    if (cpt >= 0x2B820 && cpt <= 0x2CEAF) return true;
+
+    // CJK Extension F
+    if (cpt >= 0x2CEB0 && cpt <= 0x2EBEF) return true;
+
+    // CJK Compatibility Ideographs
+    if (cpt >= 0xF900 && cpt <= 0xFAFF) return true;
+
+    // CJK Compatibility Ideographs Supplement
+    if (cpt >= 0x2F800 && cpt <= 0x2FA1F) return true;
+
+    return false;
+}
+
 std::vector<std::string> unicode_regex_split(const std::string & text, const std::vector<std::string> & regex_exprs) {
     // unicode categories
     static const std::map<std::string, int> k_ucat_enum = {

data/ext/sources/examples/talk-llama/unicode.h

@@ -4,6 +4,7 @@
 #include <string>
 #include <vector>
 
+// TODO: reimplement this structure in endian-independent way
 struct unicode_cpt_flags {
     enum {
         UNDEFINED       = 0x0001,
@@ -15,6 +16,10 @@ struct unicode_cpt_flags {
         SYMBOL          = 0x0040,  // regex: \p{S}
         CONTROL         = 0x0080,  // regex: \p{C}
         MASK_CATEGORIES = 0x00FF,
+        WHITESPACE      = 0x0100,
+        LOWERCASE       = 0x0200,
+        UPPERCASE       = 0x0400,
+        NFD             = 0x0800,
     };
 
     // codepoint type
@@ -34,11 +39,49 @@ struct unicode_cpt_flags {
 
     // decode from uint16
     inline unicode_cpt_flags(const uint16_t flags = 0) {
+#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
         *reinterpret_cast<uint16_t*>(this) = flags;
+#elif __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
+        is_undefined   = (flags & UNDEFINED)   ? 1 : 0;
+        is_number      = (flags & NUMBER)      ? 1 : 0;
+        is_letter      = (flags & LETTER)      ? 1 : 0;
+        is_separator   = (flags & SEPARATOR)   ? 1 : 0;
+        is_accent_mark = (flags & ACCENT_MARK) ? 1 : 0;
+        is_punctuation = (flags & PUNCTUATION) ? 1 : 0;
+        is_symbol      = (flags & SYMBOL)      ? 1 : 0;
+        is_control     = (flags & CONTROL)     ? 1 : 0;
+        is_whitespace  = (flags & WHITESPACE)  ? 1 : 0;
+        is_lowercase   = (flags & LOWERCASE)   ? 1 : 0;
+        is_uppercase   = (flags & UPPERCASE)   ? 1 : 0;
+        is_nfd         = (flags & NFD)         ? 1 : 0;
+#else
+#error Unexpected or undefined __BYTE_ORDER__
+#endif
     }
 
     inline uint16_t as_uint() const {
+#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
         return *reinterpret_cast<const uint16_t*>(this);
+#elif __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
+        uint16_t result =
+              is_undefined   * UNDEFINED
+            + is_number      * NUMBER
+            + is_letter      * LETTER
+            + is_separator   * SEPARATOR
+            + is_accent_mark * ACCENT_MARK
+            + is_punctuation * PUNCTUATION
+            + is_symbol      * SYMBOL
+            + is_control     * CONTROL
+            + is_whitespace  * WHITESPACE
+            + is_lowercase   * LOWERCASE
+            + is_uppercase   * UPPERCASE
+            + is_nfd         * NFD
+            ;
+
+        return result;
+#else
+#error Unexpected or undefined __BYTE_ORDER__
+#endif
     }
 
     inline uint16_t category_flag() const {
@@ -63,4 +106,6 @@ uint8_t     unicode_utf8_to_byte(const std::string & utf8);
 
 uint32_t unicode_tolower(uint32_t cpt);
 
+bool unicode_cpt_is_han(uint32_t cpt);
+
 std::vector<std::string> unicode_regex_split(const std::string & text, const std::vector<std::string> & regex_exprs);

data/ext/sources/examples/vad-speech-segments/speech.cpp

@@ -83,6 +83,8 @@ static bool vad_params_parse(int argc, char ** argv, cli_params & params) {
 static void cb_log_disable(enum ggml_log_level , const char * , void * ) { }
 
 int main(int argc, char ** argv) {
+    ggml_backend_load_all();
+
     cli_params cli_params;
 
     if (!vad_params_parse(argc, argv, cli_params)) {
@@ -109,6 +111,10 @@ int main(int argc, char ** argv) {
     struct whisper_vad_context * vctx = whisper_vad_init_from_file_with_params(
             cli_params.vad_model.c_str(),
             ctx_params);
+    if (vctx == nullptr) {
+        fprintf(stderr, "error: failed to initialize whisper context\n");
+        return 2;
+    }
 
     // Detect speech in the input audio file.
     if (!whisper_vad_detect_speech(vctx, pcmf32.data(), pcmf32.size())) {

data/ext/sources/examples/wchess/wchess.cmd/wchess.cmd.cpp

@@ -31,7 +31,7 @@ struct whisper_params {
     bool print_energy  = false;
     bool no_timestamps = true;
     bool use_gpu       = true;
-    bool flash_attn    = false;
+    bool flash_attn    = true;
 
     std::string language  = "en";
     std::string model     = "models/ggml-base.en.bin";
@@ -60,7 +60,8 @@ void whisper_print_usage(int /*argc*/, char ** argv, const whisper_params & para
     fprintf(stderr, "  -ps,        --print-special  [%-7s] print special tokens\n",                        params.print_special ? "true" : "false");
     fprintf(stderr, "  -pe,        --print-energy   [%-7s] print sound energy (for debugging)\n",          params.print_energy ? "true" : "false");
     fprintf(stderr, "  -ng,        --no-gpu         [%-7s] disable GPU\n",                                 params.use_gpu ? "false" : "true");
-    fprintf(stderr, "  -fa,        --flash-attn     [%-7s] flash attention during decoding\n",             params.flash_attn ? "true" : "false");
+    fprintf(stderr, "  -fa,        --flash-attn     [%-7s] enable flash attention during decoding\n",      params.flash_attn ? "true" : "false");
+    fprintf(stderr, "  -nfa,       --no-flash-attn  [%-7s] disable flash attention during decoding\n",     params.flash_attn ? "false" : "true");
     fprintf(stderr, "  -l LANG,    --language LANG  [%-7s] spoken language\n",                             params.language.c_str());
     fprintf(stderr, "  -m FNAME,   --model FNAME    [%-7s] model path\n",                                  params.model.c_str());
     fprintf(stderr, "  -f FNAME,   --file FNAME     [%-7s] text output file name\n",                       params.fname_out.c_str());
@@ -92,6 +93,7 @@ bool whisper_params_parse(int argc, char ** argv, whisper_params & params) {
         else if (arg == "-pe"  || arg == "--print-energy")  { params.print_energy  = true; }
         else if (arg == "-ng"  || arg == "--no-gpu")        { params.use_gpu       = false; }
         else if (arg == "-fa"  || arg == "--flash-attn")    { params.flash_attn    = true; }
+        else if (arg == "-nfa" || arg == "--no-flash-attn") { params.flash_attn    = false; }
         else if (arg == "-l"   || arg == "--language")      { params.language      = argv[++i]; }
         else if (arg == "-m"   || arg == "--model")         { params.model         = argv[++i]; }
         else if (arg == "-f"   || arg == "--file")          { params.fname_out     = argv[++i]; }
@@ -168,6 +170,8 @@ bool get_audio(std::vector<float> & pcmf32_cur) {
 }
 
 int main(int argc, char ** argv) {
+    ggml_backend_load_all();
+
     whisper_params params;
 
     if (whisper_params_parse(argc, argv, params) == false) {

data/ext/sources/examples/whisper.wasm/index-tmpl.html

@@ -52,6 +52,7 @@
                 <a href="bench.wasm/">bench</a> |
                 <a href="stream.wasm">stream</a> |
                 <a href="command.wasm/">command</a> |
+                <a href="wchess.wasm/">wchess</a> |
 
             <hr>
 
@@ -338,22 +339,22 @@
 
             function loadWhisper(model) {
                 let urls = {
-                    'tiny.en':  'https://whisper.ggerganov.com/ggml-model-whisper-tiny.en.bin',
-                    'tiny':     'https://whisper.ggerganov.com/ggml-model-whisper-tiny.bin',
-                    'base.en':  'https://whisper.ggerganov.com/ggml-model-whisper-base.en.bin',
-                    'base':     'https://whisper.ggerganov.com/ggml-model-whisper-base.bin',
-                    'small.en': 'https://whisper.ggerganov.com/ggml-model-whisper-small.en.bin',
-                    'small':    'https://whisper.ggerganov.com/ggml-model-whisper-small.bin',
-
-                    'tiny-en-q5_1':  'https://whisper.ggerganov.com/ggml-model-whisper-tiny.en-q5_1.bin',
-                    'tiny-q5_1':     'https://whisper.ggerganov.com/ggml-model-whisper-tiny-q5_1.bin',
-                    'base-en-q5_1':  'https://whisper.ggerganov.com/ggml-model-whisper-base.en-q5_1.bin',
-                    'base-q5_1':     'https://whisper.ggerganov.com/ggml-model-whisper-base-q5_1.bin',
-                    'small-en-q5_1': 'https://whisper.ggerganov.com/ggml-model-whisper-small.en-q5_1.bin',
-                    'small-q5_1':    'https://whisper.ggerganov.com/ggml-model-whisper-small-q5_1.bin',
-                    'medium-en-q5_0':'https://whisper.ggerganov.com/ggml-model-whisper-medium.en-q5_0.bin',
-                    'medium-q5_0':   'https://whisper.ggerganov.com/ggml-model-whisper-medium-q5_0.bin',
-                    'large-q5_0':    'https://whisper.ggerganov.com/ggml-model-whisper-large-q5_0.bin',
+                    'tiny.en':  'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-tiny.en.bin',
+                    'tiny':     'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-tiny.bin',
+                    'base.en':  'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.en.bin',
+                    'base':     'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.bin',
+                    'small.en': 'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-small.en.bin',
+                    'small':    'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-small.bin',
+
+                    'tiny-en-q5_1':  'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-tiny.en-q5_1.bin',
+                    'tiny-q5_1':     'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-tiny-q5_1.bin',
+                    'base-en-q5_1':  'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.en-q5_1.bin',
+                    'base-q5_1':     'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base-q5_1.bin',
+                    'small-en-q5_1': 'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-small.en-q5_1.bin',
+                    'small-q5_1':    'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-small-q5_1.bin',
+                    'medium-en-q5_0':'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-medium.en-q5_0.bin',
+                    'medium-q5_0':   'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-medium-q5_0.bin',
+                    'large-q5_0':    'https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-large-q5_0.bin',
                 };
 
                 let sizes = {