From 90dfe73f1a1b6815cfba08ec165394719005e4cf Mon Sep 17 00:00:00 2001
From: Ariver <shanghai3168@gmail.com>
Date: Sun, 07 Jun 2026 01:48:32 +0800
Subject: [PATCH] Fix X-ASR CJK spacing
---
privatevoice.src/internal/model/registry.go | 219 ++++++++++++++++++++++++++++++++++++++++++++++++++++--
1 files changed, 211 insertions(+), 8 deletions(-)
diff --git a/privatevoice.src/internal/model/registry.go b/privatevoice.src/internal/model/registry.go
index 9314941..8e30239 100644
--- a/privatevoice.src/internal/model/registry.go
+++ b/privatevoice.src/internal/model/registry.go
@@ -1,9 +1,16 @@
package model
-import "fmt"
+import (
+ "fmt"
+ "sort"
+)
const (
- BackendSenseVoice = "sensevoice"
+ BackendSenseVoice = "sensevoice"
+ BackendMoonshine = "moonshine"
+ BackendNemoTransducer = "nemo_transducer"
+ BackendQwen3ASR = "qwen3_asr"
+ BackendXASRStreaming = "x_asr_streaming"
)
var modelProfiles = map[string]ModelProfile{
@@ -23,6 +30,10 @@
},
RecommendedFor: []string{"zh-CN"},
Description: "轻量中文模型，适合中文和中英混输。",
+ DownloadURLs: []string{
+ "http://www.maikami.com/voicesnap/sensevoice.zip",
+ "https://modelscope.cn/models/sherpa-onnx/sherpa-onnx-sense-voice-zh-en-ja-ko-yue/resolve/master/sherpa-onnx-sense-voice-zh-en-ja-ko-yue-int8-2024-07-17.tar.bz2",
+ },
InstallDirName: "sensevoice-zh",
LegacyDirNames: []string{"sensevoice"},
RequiredFiles: []RequiredFileRule{
@@ -40,15 +51,191 @@
ProviderOrder: []string{"coreml", "cpu"},
NumThreads: 4,
},
+ MoonshineModelID: {
+ ID: MoonshineModelID,
+ DisplayName: "Moonshine English",
+ BackendKind: BackendMoonshine,
+ Tier: "default",
+ SupportedLanguageIDs: []string{EnglishLanguageID},
+ RecommendedFor: []string{EnglishLanguageID},
+ Description: "轻量英文离线模型，适合英文听写和英文日常输入。",
+ ApproxSize: "约 239 MiB 下载包，解压后约 270-280 MiB",
+ DownloadURLs: []string{
+ "https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-moonshine-base-en-int8.tar.bz2",
+ },
+ InstallDirName: "moonshine-en",
+ RequiredFiles: []RequiredFileRule{
+ {
+ Role: "preprocessor",
+ AllOf: []string{"preprocess.onnx"},
+ Required: true,
+ },
+ {
+ Role: "encoder",
+ AllOf: []string{"encode.int8.onnx"},
+ Required: true,
+ },
+ {
+ Role: "uncached_decoder",
+ AllOf: []string{"uncached_decode.int8.onnx"},
+ Required: true,
+ },
+ {
+ Role: "cached_decoder",
+ AllOf: []string{"cached_decode.int8.onnx"},
+ Required: true,
+ },
+ {
+ Role: "tokens",
+ AllOf: []string{"tokens.txt"},
+ Required: true,
+ },
+ },
+ ProviderOrder: []string{"cpu"},
+ NumThreads: 4,
+ },
+ ParakeetModelID: {
+ ID: ParakeetModelID,
+ DisplayName: "Parakeet English",
+ BackendKind: BackendNemoTransducer,
+ Tier: "advanced",
+ SupportedLanguageIDs: []string{EnglishLanguageID},
+ RecommendedFor: []string{EnglishLanguageID},
+ Description: "高质量英文离线模型，适合英文长句、技术词和更高准确率需求。",
+ ApproxSize: "约 478 MiB 下载包",
+ DownloadURLs: []string{
+ "https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-nemo-parakeet-unified-en-0.6b-int8-non-streaming.tar.bz2",
+ },
+ InstallDirName: "parakeet-en",
+ RequiredFiles: []RequiredFileRule{
+ {
+ Role: "encoder",
+ AllOf: []string{"encoder.int8.onnx"},
+ Required: true,
+ },
+ {
+ Role: "decoder",
+ AllOf: []string{"decoder.int8.onnx"},
+ Required: true,
+ },
+ {
+ Role: "joiner",
+ AllOf: []string{"joiner.int8.onnx"},
+ Required: true,
+ },
+ {
+ Role: "tokens",
+ AllOf: []string{"tokens.txt"},
+ Required: true,
+ },
+ },
+ ProviderOrder: []string{"cpu"},
+ NumThreads: 4,
+ },
+ Qwen3ASRModelID: {
+ ID: Qwen3ASRModelID,
+ DisplayName: "Qwen3-ASR",
+ BackendKind: BackendQwen3ASR,
+ Tier: "advanced",
+ SupportedLanguageIDs: []string{DefaultLanguageID, "zh-Hans", "zh-Hant", "yue"},
+ RecommendedFor: []string{DefaultLanguageID},
+ Description: "高质量中文离线模型，适合中文长句、技术词和中英混输准确率需求。",
+ ApproxSize: "约 838 MiB 下载包，解压后约 1.9 GiB",
+ DownloadURLs: []string{
+ "https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-qwen3-asr-0.6B-int8-2026-03-25.tar.bz2",
+ },
+ InstallDirName: "qwen3-asr-0.6b",
+ RequiredFiles: []RequiredFileRule{
+ {
+ Role: "conv_frontend",
+ AllOf: []string{"conv_frontend.onnx"},
+ Required: true,
+ },
+ {
+ Role: "encoder",
+ AllOf: []string{"encoder.int8.onnx"},
+ Required: true,
+ },
+ {
+ Role: "decoder",
+ AllOf: []string{"decoder.int8.onnx"},
+ Required: true,
+ },
+ {
+ Role: "tokenizer",
+ AllOf: []string{"tokenizer"},
+ Required: true,
+ Directory: true,
+ },
+ {
+ Role: "tokenizer_merges",
+ AllOf: []string{"tokenizer/merges.txt"},
+ Required: true,
+ },
+ {
+ Role: "tokenizer_vocab",
+ AllOf: []string{"tokenizer/vocab.json"},
+ Required: true,
+ },
+ },
+ ProviderOrder: []string{"cpu"},
+ NumThreads: 4,
+ },
+ XASRZhEn960ModelID: {
+ ID: XASRZhEn960ModelID,
+ DisplayName: "X-ASR zh-en 960ms",
+ BackendKind: BackendXASRStreaming,
+ Tier: "experimental",
+ SupportedLanguageIDs: []string{DefaultLanguageID, "zh-Hans", "zh-Hant", "yue", EnglishLanguageID},
+ RecommendedFor: []string{DefaultLanguageID, EnglishLanguageID},
+ Description: "实验性中英流式模型，适合评估中英混输和未来边说边出字体验。",
+ ApproxSize: "单个 960 ms 模型目录约 586 MiB",
+ InstallDirName: "x-asr-zh-en-960ms",
+ RequiredFiles: []RequiredFileRule{
+ {
+ Role: "encoder",
+ AllOf: []string{"encoder-960ms.onnx"},
+ Required: true,
+ },
+ {
+ Role: "decoder",
+ AllOf: []string{"decoder-960ms.onnx"},
+ Required: true,
+ },
+ {
+ Role: "joiner",
+ AllOf: []string{"joiner-960ms.onnx"},
+ Required: true,
+ },
+ {
+ Role: "tokens",
+ AllOf: []string{"tokens.txt"},
+ Required: true,
+ },
+ },
+ ProviderOrder: []string{"cpu"},
+ NumThreads: 1,
+ },
}
var languageProfiles = map[string]LanguageProfile{
- "zh-CN": {
- ID: "zh-CN",
- DisplayName: "中文",
- UILocale: "zh-CN",
- SystemMatchers: []string{"zh", "zh-CN", "zh-Hans", "zh-Hant"},
- DefaultModelID: DefaultModelID,
+ DefaultLanguageID: {
+ ID: DefaultLanguageID,
+ DisplayName: "Chinese",
+ NativeName: "中文",
+ UILocale: "zh",
+ SystemMatchers: []string{"zh", "zh-CN", "zh-Hans", "zh-Hant"},
+ DefaultModelID: DefaultModelID,
+ UpgradeModelIDs: []string{Qwen3ASRModelID, XASRZhEn960ModelID},
+ },
+ EnglishLanguageID: {
+ ID: EnglishLanguageID,
+ DisplayName: "English",
+ NativeName: "English",
+ UILocale: "en",
+ SystemMatchers: []string{"en", "en-US", "en-GB", "en-AU", "en-CA"},
+ DefaultModelID: MoonshineModelID,
+ UpgradeModelIDs: []string{ParakeetModelID, XASRZhEn960ModelID},
},
}
@@ -95,5 +282,21 @@
for _, profile := range languageProfiles {
profiles = append(profiles, profile)
}
+ sort.Slice(profiles, func(i, j int) bool {
+ if profiles[i].ID == DefaultLanguageID {
+ return true
+ }
+ if profiles[j].ID == DefaultLanguageID {
+ return false
+ }
+ return profiles[i].ID < profiles[j].ID
+ })
return profiles
}
+
+func NormalizeLanguageID(id string) string {
+ if _, err := GetLanguageProfile(id); err != nil {
+ return DefaultLanguageID
+ }
+ return id
+}
--
Gitblit v1.9.3