diff --git a/img/VRCT_now_downloading.png b/img/VRCT_now_downloading.png index 5e5b9daa..8e372d87 100644 Binary files a/img/VRCT_now_downloading.png and b/img/VRCT_now_downloading.png differ diff --git a/locales/en.yml b/locales/en.yml index f68aa32c..93f07195 100644 --- a/locales/en.yml +++ b/locales/en.yml @@ -79,7 +79,7 @@ config_window: transcription: Transcription transcription_mic: Mic transcription_speaker: Speaker - transcription_internal_model: Internal Model + transcription_internal_model: Transcription Model others: Others others_send_message_formats: Message Formats (Send) others_received_message_formats: Message Formats (Received) @@ -126,21 +126,6 @@ config_window: small: "Basic model (%{capacity})" large: "High accuracy model (%{capacity})" - use_whisper_feature: - label: Use Whisper Feature - desc: Description - - whisper_weight_type: - label: Select Whisper Model - desc: Description - tiny: "tiny model (%{capacity})" - base: "base model (%{capacity})" - small: "small model (%{capacity})" - medium: "medium model (%{capacity})" - large_v1: "large_v1 model (%{capacity})" - large_v2: "large_v2 model (%{capacity})" - large_v3: "large_v3 model (%{capacity})" - deepl_auth_key: label: DeepL Auth Key desc: Please select %{translator} on the main screen with DeepL_API when using. ※Some languages may not be supported. @@ -203,6 +188,15 @@ config_window: desc: It is the lower limit for the number of transcribed words, and only when this number is exceeded will the transcription results be displayed logs. error_message: You can set a number equal to or greater than 0. + use_whisper_feature: + label: Use Whisper Model As Transcription + desc: In some languages, the accuracy of speech recognition may improve. During speech recognition usage, CPU usage increases, so please consider your PC specs before using this feature. + + whisper_weight_type: + label: Select Whisper Model + desc: Generally, models with larger capacity tend to have higher accuracy, but this also results in longer transcription times and increased CPU usage. Please refer to the documentation for explanations of each model. + model_template: "%{model_name} model (%{capacity})" + recommended_model_template: "%{model_name} model (%{capacity}) (Recommended)" auto_clear_the_message_box: label: Auto Clear The Message Box diff --git a/locales/ja.yml b/locales/ja.yml index 6bedafbd..5d0b6acf 100644 --- a/locales/ja.yml +++ b/locales/ja.yml @@ -79,6 +79,7 @@ config_window: transcription: 音声認識 transcription_mic: マイク transcription_speaker: スピーカー + transcription_internal_model: 音声認識モデル others: その他 advanced_settings: 高度な設定 @@ -184,6 +185,16 @@ config_window: desc: 文字起こしされた単語数の下限値で、この数値を超えた場合のみ結果をログに表示します。 error_message: 0以上の数値を設定できます。 + use_whisper_feature: + label: 音声認識にWhisperモデルを使用する + desc: 一部の言語では、音声認識の精度が向上するかもしれません。音声認識使用中、CPUの使用率が上がるので、お使いのPCスペックと相談してこの機能を使用してください。 + + whisper_weight_type: + label: Whisperモデルのタイプ + desc: 基本的に、容量が多いモデルほど精度は高いですが、文字起こしまでの時間が伸び、CPU使用率も増加します。各モデルの説明はドキュメントをご覧ください。 + model_template: "%{model_name} モデル (%{capacity})" + recommended_model_template: "%{model_name} モデル (%{capacity}) (推奨)" + auto_clear_the_message_box: label: 送信後はチャットボックスを空にする diff --git a/view.py b/view.py index 91074d93..2d9b8fb0 100644 --- a/view.py +++ b/view.py @@ -913,8 +913,8 @@ class View(): @staticmethod def getSelectableCtranslate2WeightTypeDict(): return { - config._SELECTABLE_CTRANSLATE2_WEIGHT_TYPE_DICT["Small"]: i18n.t("config_window.ctranslate2_weight_type.small", capacity="418MB"), - config._SELECTABLE_CTRANSLATE2_WEIGHT_TYPE_DICT["Large"]: i18n.t("config_window.ctranslate2_weight_type.large", capacity="1.2GB"), + config.SELECTABLE_CTRANSLATE2_WEIGHT_TYPE_DICT["Small"]: i18n.t("config_window.ctranslate2_weight_type.small", capacity="418MB"), + config.SELECTABLE_CTRANSLATE2_WEIGHT_TYPE_DICT["Large"]: i18n.t("config_window.ctranslate2_weight_type.large", capacity="1.2GB"), } def useTranslationFeatureProcess(self, state:str): @@ -952,14 +952,21 @@ class View(): @staticmethod def getSelectableWhisperWeightTypeDict(): + def callI18n(model_name, capacity, is_recommended=False): + if is_recommended is True: + return i18n.t("config_window.whisper_weight_type.recommended_model_template", model_name=model_name, capacity=capacity) + else: + return i18n.t("config_window.whisper_weight_type.model_template", model_name=model_name, capacity=capacity) + + DICT_DATA = config.SELECTABLE_WHISPER_WEIGHT_TYPE_DICT return { - config.SELECTABLE_WHISPER_WEIGHT_TYPE_DICT["tiny"]: i18n.t("config_window.whisper_weight_type.tiny", capacity="74.5MB"), - config.SELECTABLE_WHISPER_WEIGHT_TYPE_DICT["base"]: i18n.t("config_window.whisper_weight_type.base", capacity="141MB"), - config.SELECTABLE_WHISPER_WEIGHT_TYPE_DICT["small"]: i18n.t("config_window.whisper_weight_type.small", capacity="463MB"), - config.SELECTABLE_WHISPER_WEIGHT_TYPE_DICT["medium"]: i18n.t("config_window.whisper_weight_type.medium", capacity="1.42GB"), - config.SELECTABLE_WHISPER_WEIGHT_TYPE_DICT["large-v1"]: i18n.t("config_window.whisper_weight_type.large_v1", capacity="2.87GB"), - config.SELECTABLE_WHISPER_WEIGHT_TYPE_DICT["large-v2"]: i18n.t("config_window.whisper_weight_type.large_v2", capacity="2.87GB"), - config.SELECTABLE_WHISPER_WEIGHT_TYPE_DICT["large-v3"]: i18n.t("config_window.whisper_weight_type.large_v3", capacity="2.87GB"), + DICT_DATA["tiny"]: callI18n("tiny", "74.5MB"), + DICT_DATA["base"]: callI18n("base", "141MB", True), + DICT_DATA["small"]: callI18n("small", "463MB"), + DICT_DATA["medium"]: callI18n("medium", "1.42GB"), + DICT_DATA["large-v1"]: callI18n("large-v1", "2.87GB"), + DICT_DATA["large-v2"]: callI18n("large-v2", "2.87GB"), + DICT_DATA["large-v3"]: callI18n("large-v3", "2.87GB"), } # Open Webpage Functions