whispercpp 1.3.2 → 1.3.3

data/ext/sources/ggml/src/gguf.cpp

@@ -335,7 +335,11 @@ struct gguf_context * gguf_init_from_file_impl(FILE * file, struct gguf_init_par
 
         for (uint32_t i = 0; i < magic.size(); i++) {
             if (magic[i] != GGUF_MAGIC[i]) {
-                GGML_LOG_ERROR("%s: invalid magic characters: '%c%c%c%c', expected 'GGUF'\n", __func__, magic[0], magic[1], magic[2], magic[3]);
+                char c0 = isprint(magic[0]) ? magic[0] : '?';
+                char c1 = isprint(magic[1]) ? magic[1] : '?';
+                char c2 = isprint(magic[2]) ? magic[2] : '?';
+                char c3 = isprint(magic[3]) ? magic[3] : '?';
+                GGML_LOG_ERROR("%s: invalid magic characters: '%c%c%c%c', expected 'GGUF'\n", __func__, c0, c1, c2, c3);
                 gguf_free(ctx);
                 return nullptr;
             }
@@ -347,11 +351,28 @@ struct gguf_context * gguf_init_from_file_impl(FILE * file, struct gguf_init_par
     int64_t n_tensors = 0;
 
     if (ok && gr.read(ctx->version)) {
-        if (ctx->version == 1) {
+        if (ok && ctx->version == 0) {
+            GGML_LOG_ERROR("%s: bad GGUF version: %" PRIu32 "\n", __func__, ctx->version);
+            ok = false;
+        }
+
+        /*
+         * bit layout is different when reading non-native endian models.
+         * assuming that the GGUF version is 3, the non-native endian model
+         * would read it as 0x30000000. we can use the AND operation against
+         * the last 4 hexadecimal digits to check if the model is the same
+         * endianness as the host system.
+        */
+        if (ok && (ctx->version & 0x0000FFFF) == 0x00000000) {
+            GGML_LOG_ERROR("%s: failed to load model: this GGUF file version %" PRIu32 " is extremely large, is there a mismatch between the host and model endianness?\n", __func__, ctx->version);
+            ok = false;
+        }
+
+        if (ok && ctx->version == 1) {
             GGML_LOG_ERROR("%s: GGUFv1 is no longer supported, please use a more up-to-date version\n", __func__);
             ok = false;
         }
-        if (ctx->version > GGUF_VERSION) {
+        if (ok && ctx->version > GGUF_VERSION) {
             GGML_LOG_ERROR("%s: this GGUF file is version %" PRIu32 " but this software only supports up to version %d\n",
                 __func__, ctx->version, GGUF_VERSION);
             ok = false;

data/ext/sources/include/whisper.h

@@ -198,6 +198,8 @@ extern "C" {
         float samples_overlap;         // Overlap in seconds when copying audio samples from speech segment.
     } whisper_vad_params;
 
+    WHISPER_API const char * whisper_version(void);
+
     // Various functions for loading a ggml whisper model.
     // Allocate (almost) all memory needed for the model.
     // Return NULL on failure

data/ext/sources/src/CMakeLists.txt

@@ -58,6 +58,7 @@ if (WHISPER_COREML)
     set(TARGET whisper.coreml)
 
     add_library(${TARGET}
+        coreml/whisper-compat.m
         coreml/whisper-encoder.h
         coreml/whisper-encoder.mm
         coreml/whisper-encoder-impl.h
@@ -76,6 +77,7 @@ if (WHISPER_COREML)
         COMPILE_FLAGS "-fobjc-arc"
         XCODE_ATTRIBUTE_CLANG_ENABLE_OBJC_ARC YES
         )
+
     set_target_properties(${TARGET} PROPERTIES FOLDER "libs")
 endif()
 

data/ext/sources/src/coreml/whisper-compat.h

@@ -0,0 +1,10 @@
+#import <CoreML/CoreML.h>
+
+@interface MLModel (Compat)
+- (void) predictionFromFeatures:(id<MLFeatureProvider>) input
+              completionHandler:(void (^)(id<MLFeatureProvider> output, NSError * error)) completionHandler;
+
+- (void) predictionFromFeatures:(id<MLFeatureProvider>) input
+                        options:(MLPredictionOptions *) options
+              completionHandler:(void (^)(id<MLFeatureProvider> output, NSError * error)) completionHandler;
+@end

data/ext/sources/src/coreml/whisper-compat.m

@@ -0,0 +1,35 @@
+#import "whisper-compat.h"
+#import <Foundation/Foundation.h>
+
+@implementation MLModel (Compat)
+
+#if !defined(MAC_OS_X_VERSION_14_00) || MAC_OS_X_VERSION_MAX_ALLOWED < MAC_OS_X_VERSION_14_00
+
+- (void) predictionFromFeatures:(id<MLFeatureProvider>) input
+              completionHandler:(void (^)(id<MLFeatureProvider> output, NSError * error)) completionHandler {
+    [NSOperationQueue.new addOperationWithBlock:^{
+        NSError *error = nil;
+        id<MLFeatureProvider> prediction = [self predictionFromFeatures:input error:&error];
+
+        [NSOperationQueue.mainQueue addOperationWithBlock:^{
+            completionHandler(prediction, error);
+        }];
+    }];
+}
+
+- (void) predictionFromFeatures:(id<MLFeatureProvider>) input
+                        options:(MLPredictionOptions *) options
+              completionHandler:(void (^)(id<MLFeatureProvider> output, NSError * error)) completionHandler {
+    [NSOperationQueue.new addOperationWithBlock:^{
+        NSError *error = nil;
+        id<MLFeatureProvider> prediction = [self predictionFromFeatures:input options:options error:&error];
+
+        [NSOperationQueue.mainQueue addOperationWithBlock:^{
+            completionHandler(prediction, error);
+        }];
+    }];
+}
+
+#endif
+
+@end

data/ext/sources/src/coreml/whisper-decoder-impl.m

@@ -8,6 +8,7 @@
 #error This file must be compiled with automatic reference counting enabled (-fobjc-arc)
 #endif
 
+#import "whisper-compat.h"
 #import "whisper-decoder-impl.h"
 
 @implementation whisper_decoder_implInput

data/ext/sources/src/coreml/whisper-encoder-impl.m

@@ -8,6 +8,7 @@
 #error This file must be compiled with automatic reference counting enabled (-fobjc-arc)
 #endif
 
+#import "whisper-compat.h"
 #import "whisper-encoder-impl.h"
 
 @implementation whisper_encoder_implInput