Skip to content

🌐 AI-Powered i18n Synchronization #421

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
DearVa merged 1 commit into from
Jun 10, 2026
Merged

🌐 AI-Powered i18n Synchronization #421

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
21 changes: 18 additions & 3 deletions src/Everywhere.Core/I18N/Strings.de.resx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,6 +94,12 @@
<data name="ShortcutSettings_ChatWindow_Desription" xml:space="preserve">
<value>Legen Sie eine Tastenkombination fest, um das Chatfenster jederzeit aufzurufen.</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_Temperature_Description" xml:space="preserve">
<value>Höhere Werte (z. B. 0,8) machen die Ausgabe zufälliger, während niedrigere Werte (z. B. 0,2) gezielter und bestimmter sind. Wir empfehlen normalerweise, diesen Parameter oder den top_p-Parameter anzupassen, jedoch nicht beide gleichzeitig.</value>
</data>
<data name="Assistant_TopP_Header" xml:space="preserve">
<value>top_p</value>
</data>
Expand Down Expand Up @@ -2859,16 +2865,25 @@ Veröffentlichungsdatum: {2}</value>
<data name="CodeBlock_ToggleTextWrap_Tip" xml:space="preserve">
<value>Textumbruch umschalten</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_PresencePenalty_Header" xml:space="preserve">
<value>presence_penalty</value>
</data>
<data name="Assistant_PresencePenalty_Description" xml:space="preserve">
<value>Ein positiver Wert bestraft neue Wörter basierend darauf, ob sie im aktuellen Text erscheinen, wodurch die Wahrscheinlichkeit erhöht wird, dass das Modell neue Themen diskutiert.</value>
</data>
<data name="Assistant_TopP_Description" xml:space="preserve">
<value>Eine alternative Methode zur Temperatursampling wird als „Kernsampling“ bezeichnet, bei der das Modell nur die Token berücksichtigt, die in der Wahrscheinlichkeitssumme der obersten top_p liegen. Daher bedeutet 0,1, dass nur die Token in den obersten 10 % der Wahrscheinlichkeitssumme berücksichtigt werden.</value>
</data>
<data name="Assistant_FrequencyPenalty_Header" xml:space="preserve">
<value>frequency_penalty</value>
</data>
<data name="Assistant_FrequencyPenalty_Description" xml:space="preserve">
<value>Ein positiver Wert bestraft neue Wörter basierend auf ihrer Häufigkeit im aktuellen Text, wodurch die Wahrscheinlichkeit verringert wird, dass das Modell dieselbe Zeile wiederholt.</value>
</data>
<data name="Assistant_TopK_Header" xml:space="preserve">
<value>top_k</value>
</data>
<data name="Assistant_TopK_Description" xml:space="preserve">
<value>Für jedes nachfolgende Token wird nur aus den obersten K Optionen ausgewählt.</value>
</data>
</root>
21 changes: 18 additions & 3 deletions src/Everywhere.Core/I18N/Strings.es.resx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,6 +94,12 @@
<data name="ShortcutSettings_ChatWindow_Desription" xml:space="preserve">
<value>Establece un atajo para invocar la ventana de chat en cualquier momento.</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_Temperature_Description" xml:space="preserve">
<value>Un valor más alto (como 0.8) hará que los resultados sean más aleatorios, mientras que un valor más bajo (como 0.2) los hará más específicos y deterministas. Normalmente recomendamos ajustar este parámetro o el parámetro top_p, pero no ambos al mismo tiempo.</value>
</data>
<data name="Assistant_TopP_Header" xml:space="preserve">
<value>top_p</value>
</data>
Expand Down Expand Up @@ -2860,16 +2866,25 @@ Fecha de lanzamiento: {2}</value>
<data name="CodeBlock_ToggleTextWrap_Tip" xml:space="preserve">
<value>Alternar ajuste de texto</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_PresencePenalty_Header" xml:space="preserve">
<value>presence_penalty</value>
</data>
<data name="Assistant_PresencePenalty_Description" xml:space="preserve">
<value>Un valor positivo penaliza las palabras nuevas según su aparición en el texto actual, aumentando la probabilidad de que el modelo discuta nuevos temas.</value>
</data>
<data name="Assistant_TopP_Description" xml:space="preserve">
<value>Un método alternativo de muestreo de temperatura se llama 'muestreo de núcleo', donde el modelo solo considera los tokens con la masa de probabilidad en el top_p. Por lo tanto, 0.1 significa que solo se consideran los tokens en el top 10% de la masa de probabilidad.</value>
</data>
<data name="Assistant_FrequencyPenalty_Header" xml:space="preserve">
<value>frequency_penalty</value>
</data>
<data name="Assistant_FrequencyPenalty_Description" xml:space="preserve">
<value>Un valor positivo penaliza las palabras nuevas según su frecuencia de aparición en el texto actual, reduciendo la probabilidad de que el modelo repita la misma línea sin cambios.</value>
</data>
<data name="Assistant_TopK_Header" xml:space="preserve">
<value>top_k</value>
</data>
<data name="Assistant_TopK_Description" xml:space="preserve">
<value>Para cada token subsiguiente, se muestrea solo de las opciones en el top K.</value>
</data>
</root>
21 changes: 18 additions & 3 deletions src/Everywhere.Core/I18N/Strings.fr.resx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,6 +94,12 @@
<data name="ShortcutSettings_ChatWindow_Desription" xml:space="preserve">
<value>Définissez un raccourci pour invoquer la fenêtre de chat à tout moment.</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_Temperature_Description" xml:space="preserve">
<value>Une valeur plus élevée (comme 0.8) rendra les résultats plus aléatoires, tandis qu'une valeur plus basse (comme 0.2) les rendra plus ciblés et déterministes. Nous recommandons généralement d'ajuster ce paramètre ou le paramètre top_p, mais pas les deux en même temps.</value>
</data>
<data name="Assistant_TopP_Header" xml:space="preserve">
<value>top_p</value>
</data>
Expand Down Expand Up @@ -2860,16 +2866,25 @@ Date de publication : {2}</value>
<data name="CodeBlock_ToggleTextWrap_Tip" xml:space="preserve">
<value>Basculer le retour à la ligne du texte</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_PresencePenalty_Header" xml:space="preserve">
<value>presence_penalty</value>
</data>
<data name="Assistant_PresencePenalty_Description" xml:space="preserve">
<value>Une valeur positive pénalise les nouveaux mots en fonction de leur présence dans le texte actuel, augmentant ainsi la probabilité que le modèle aborde de nouveaux sujets.</value>
</data>
<data name="Assistant_TopP_Description" xml:space="preserve">
<value>Une méthode alternative d'échantillonnage de température est appelée « échantillonnage nucleus », où le modèle ne considère que les jetons dont la masse de probabilité est dans le top_p. Ainsi, 0.1 signifie que seuls les jetons dans le top 10 % de la masse de probabilité sont pris en compte.</value>
</data>
<data name="Assistant_FrequencyPenalty_Header" xml:space="preserve">
<value>frequency_penalty</value>
</data>
<data name="Assistant_FrequencyPenalty_Description" xml:space="preserve">
<value>Une valeur positive pénalise les nouveaux mots en fonction de leur fréquence d'apparition dans le texte actuel, réduisant ainsi la probabilité que le modèle répète la même ligne.</value>
</data>
<data name="Assistant_TopK_Header" xml:space="preserve">
<value>top_k</value>
</data>
<data name="Assistant_TopK_Description" xml:space="preserve">
<value>Pour chaque jeton suivant, l'échantillonnage se fait uniquement parmi les K meilleures options.</value>
</data>
</root>
21 changes: 18 additions & 3 deletions src/Everywhere.Core/I18N/Strings.it.resx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,6 +94,12 @@
<data name="ShortcutSettings_ChatWindow_Desription" xml:space="preserve">
<value>Imposta una scorciatoia per richiamare la finestra di chat in qualsiasi momento.</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_Temperature_Description" xml:space="preserve">
<value>Valori più alti (come 0.8) rendono l'output più casuale, mentre valori più bassi (come 0.2) lo rendono più mirato e certo. Si consiglia di regolare questo parametro o il parametro top_p, ma non entrambi contemporaneamente.</value>
</data>
<data name="Assistant_TopP_Header" xml:space="preserve">
<value>top_p</value>
</data>
Expand Down Expand Up @@ -2860,16 +2866,25 @@ Data di rilascio: {2}</value>
<data name="CodeBlock_ToggleTextWrap_Tip" xml:space="preserve">
<value>Attiva/disattiva il ritorno a capo</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_PresencePenalty_Header" xml:space="preserve">
<value>presence_penalty</value>
</data>
<data name="Assistant_PresencePenalty_Description" xml:space="preserve">
<value>Un valore positivo penalizza le nuove parole in base alla loro presenza nel testo corrente, aumentando la probabilità che il modello discuta nuovi argomenti.</value>
</data>
<data name="Assistant_TopP_Description" xml:space="preserve">
<value>Un metodo alternativo di campionamento della temperatura è chiamato 'campionamento nucleare', in cui il modello considera solo i token con probabilità cumulative superiori a top_p. Quindi, 0.1 significa considerare solo i token con le prime 10% di probabilità.</value>
</data>
<data name="Assistant_FrequencyPenalty_Header" xml:space="preserve">
<value>frequency_penalty</value>
</data>
<data name="Assistant_FrequencyPenalty_Description" xml:space="preserve">
<value>Un valore positivo penalizza le nuove parole in base alla loro frequenza nel testo corrente, riducendo la probabilità che il modello ripeta la stessa riga.</value>
</data>
<data name="Assistant_TopK_Header" xml:space="preserve">
<value>top_k</value>
</data>
<data name="Assistant_TopK_Description" xml:space="preserve">
<value>Per ogni token successivo, si campiona solo dalle prime K opzioni.</value>
</data>
</root>
21 changes: 18 additions & 3 deletions src/Everywhere.Core/I18N/Strings.ja.resx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,6 +94,12 @@
<data name="ShortcutSettings_ChatWindow_Desription" xml:space="preserve">
<value>いつでもチャットウィンドウを呼び出すためのショートカットを設定します。</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_Temperature_Description" xml:space="preserve">
<value>高い値（例：0.8）は出力をよりランダムにし、低い値（例：0.2）はより特定的で確定的にします。このパラメータまたは top_p パラメータを調整することをお勧めしますが、両方を同時に調整しないでください。</value>
</data>
<data name="Assistant_TopP_Header" xml:space="preserve">
<value>top_p</value>
</data>
Expand Down Expand Up @@ -2860,16 +2866,25 @@
<data name="CodeBlock_ToggleTextWrap_Tip" xml:space="preserve">
<value>テキストの折り返しを切り替え</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_PresencePenalty_Header" xml:space="preserve">
<value>presence_penalty</value>
</data>
<data name="Assistant_PresencePenalty_Description" xml:space="preserve">
<value>正の値は新しい単語が現在のテキストに出現するかどうかに基づいて罰則を与え、新しいテーマについてのモデルの議論の可能性を高めます。</value>
</data>
<data name="Assistant_TopP_Description" xml:space="preserve">
<value>温度サンプリングの代替手法は「コアサンプリング」と呼ばれ、モデルは確率質量が上位 top_p のトークンのみを考慮します。したがって、0.1 は確率質量が上位 10% のトークンのみを考慮することを意味します。</value>
</data>
<data name="Assistant_FrequencyPenalty_Header" xml:space="preserve">
<value>frequency_penalty</value>
</data>
<data name="Assistant_FrequencyPenalty_Description" xml:space="preserve">
<value>正の値は新しい単語の現在のテキスト内での出現頻度に基づいて罰則を与え、モデルが同じ行を繰り返す可能性を減らします。</value>
</data>
<data name="Assistant_TopK_Header" xml:space="preserve">
<value>top_k</value>
</data>
<data name="Assistant_TopK_Description" xml:space="preserve">
<value>次の各トークンについて、上位 K のオプションからのみサンプリングします。</value>
</data>
</root>
21 changes: 18 additions & 3 deletions src/Everywhere.Core/I18N/Strings.ko.resx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,6 +94,12 @@
<data name="ShortcutSettings_ChatWindow_Desription" xml:space="preserve">
<value>언제든지 채팅 창을 호출할 수 있는 단축키를 설정합니다.</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_Temperature_Description" xml:space="preserve">
<value>높은 값(예: 0.8)은 출력 결과를 더 무작위적으로 만들고, 낮은 값(예: 0.2)은 더 집중적이고 확실하게 만듭니다. 일반적으로 이 매개변수나 top_p 매개변수를 조정할 것을 권장하지만 두 가지를 동시에 조정하지는 마세요.</value>
</data>
<data name="Assistant_TopP_Header" xml:space="preserve">
<value>top_p</value>
</data>
Expand Down Expand Up @@ -2860,16 +2866,25 @@
<data name="CodeBlock_ToggleTextWrap_Tip" xml:space="preserve">
<value>텍스트 줄 바꿈 전환</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_PresencePenalty_Header" xml:space="preserve">
<value>presence_penalty</value>
</data>
<data name="Assistant_PresencePenalty_Description" xml:space="preserve">
<value>양수 값은 현재 텍스트에 새로운 단어가 나타나는지 여부에 따라 벌점을 부여하여 모델이 새로운 주제를 논의할 가능성을 높입니다.</value>
</data>
<data name="Assistant_TopP_Description" xml:space="preserve">
<value>온도 샘플링의 대안 방법인 '핵심 샘플링'에서는 모델이 확률 질량이 상위 top_p에 해당하는 토큰만 고려합니다. 따라서 0.1은 확률 질량이 상위 10%에 해당하는 토큰만 고려함을 의미합니다.</value>
</data>
<data name="Assistant_FrequencyPenalty_Header" xml:space="preserve">
<value>frequency_penalty</value>
</data>
<data name="Assistant_FrequencyPenalty_Description" xml:space="preserve">
<value>양수 값은 현재 텍스트에서 새로운 단어의 출현 빈도에 따라 벌점을 부여하여 모델이 동일한 줄을 반복할 확률을 낮춥니다.</value>
</data>
<data name="Assistant_TopK_Header" xml:space="preserve">
<value>top_k</value>
</data>
<data name="Assistant_TopK_Description" xml:space="preserve">
<value>후속 각 토큰에 대해 상위 K의 옵션 중에서만 샘플링합니다.</value>
</data>
</root>
21 changes: 18 additions & 3 deletions src/Everywhere.Core/I18N/Strings.pt-br.resx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,6 +94,12 @@
<data name="ShortcutSettings_ChatWindow_Desription" xml:space="preserve">
<value>Defina um atalho para abrir a janela de chat a qualquer momento.</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_Temperature_Description" xml:space="preserve">
<value>Valores mais altos (como 0.8) tornam os resultados mais aleatórios, enquanto valores mais baixos (como 0.2) tornam-nos mais focados e certos. Geralmente, recomendamos ajustar este parâmetro ou o parâmetro top_p, mas não ambos ao mesmo tempo.</value>
</data>
<data name="Assistant_TopP_Header" xml:space="preserve">
<value>top_p</value>
</data>
Expand Down Expand Up @@ -2861,16 +2867,25 @@ Data de lançamento: {2}</value>
<data name="CodeBlock_ToggleTextWrap_Tip" xml:space="preserve">
<value>Alternar quebra de texto</value>
</data>
<data name="Assistant_Temperature_Header" xml:space="preserve">
<value>temperature</value>
</data>
<data name="Assistant_PresencePenalty_Header" xml:space="preserve">
<value>presence_penalty</value>
</data>
<data name="Assistant_PresencePenalty_Description" xml:space="preserve">
<value>Valores positivos penalizam novas palavras com base em sua presença no texto atual, aumentando a probabilidade do modelo discutir novos tópicos.</value>
</data>
<data name="Assistant_TopP_Description" xml:space="preserve">
<value>Um método alternativo de amostragem de temperatura é chamado de 'amostragem de núcleo', onde o modelo considera apenas os tokens que estão entre os top_p em termos de massa de probabilidade. Assim, 0.1 significa considerar apenas os tokens que estão entre os 10% mais altos.</value>
</data>
<data name="Assistant_FrequencyPenalty_Header" xml:space="preserve">
<value>frequency_penalty</value>
</data>
<data name="Assistant_FrequencyPenalty_Description" xml:space="preserve">
<value>Valores positivos penalizam novas palavras com base na frequência de sua ocorrência no texto atual, reduzindo a probabilidade do modelo repetir a mesma linha.</value>
</data>
<data name="Assistant_TopK_Header" xml:space="preserve">
<value>top_k</value>
</data>
<data name="Assistant_TopK_Description" xml:space="preserve">
<value>Para cada token subsequente, a amostragem é feita apenas entre as opções classificadas entre os K melhores.</value>
</data>
</root>
Loading
Loading