injectlib/WeKnora
1
0
Fork 0
mirror of https://github.com/Tencent/WeKnora.git synced 2025-11-25 03:15:00 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix: Distinguish between Embedding and Chat detection methods

Browse Source
This commit is contained in:
wizardchen
2025-09-08 21:03:22 +08:00
committed by lyingbug
parent 277607b92f
commit 46b3de2b24
2 changed files with 77 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files

76
frontend/src/views/initialization/InitializationConfig.vue
View File

@@ -282,7 +282,7 @@
name="refresh"
class="refresh-icon"
:class="{ spinning: modelStatus.embedding.checking }"
@click="checkRemoteModelStatus('embedding')"
@click="checkEmbeddingModelStatus()"
/>
</t-tooltip>
<t-icon
@@ -1422,7 +1422,7 @@ const onModelNameInput = (type: 'llm' | 'embedding' | 'vlm') => {
if (type === 'llm' && formData.llm.source === 'remote' && formData.llm.baseUrl) {
await checkRemoteModelStatus('llm');
} else if (type === 'embedding' && formData.embedding.source === 'remote' && formData.embedding.baseUrl) {
await checkRemoteModelStatus('embedding');
await checkEmbeddingModelStatus();
} else if (type === 'vlm' && !isVlmOllama.value) {
// VLM远程API校验可以在这里添加
}
@@ -1465,7 +1465,11 @@ const onRemoteConfigChange = async (type: 'llm' | 'embedding') => {
// 如果配置完整，检查模型
if (formData[type].modelName && formData[type].baseUrl) {
await checkRemoteModelStatus(type);
if (type === 'llm') {
await checkRemoteModelStatus(type);
} else if (type === 'embedding') {
await checkEmbeddingModelStatus();
}
}
};
@@ -1492,13 +1496,17 @@ const onRemoteConfigInput = async (type: 'llm' | 'embedding') => {
form.value?.validate([`${type}.modelName`, `${type}.baseUrl`]);
// 自动检查远程API模型状态
await checkRemoteModelStatus(type);
if (type === 'llm') {
await checkRemoteModelStatus(type);
} else if (type === 'embedding') {
await checkEmbeddingModelStatus();
}
}
}, 500); // 500ms防抖延迟
};
// 检查远程模型
const checkRemoteModelStatus = async (type: 'llm' | 'embedding') => {
const checkRemoteModelStatus = async (type: 'llm') => {
if (!formData[type].modelName || !formData[type].baseUrl) {
return;
}
@@ -1752,15 +1760,8 @@ onMounted(async () => {
// 加载当前配置
await loadCurrentConfig();
// 检查Ollama状态
const needOllamaCheck =
formData.llm.source === 'local' ||
formData.embedding.source === 'local' ||
(formData.multimodal.enabled && formData.multimodal.vlm.interfaceType === 'ollama');
if (needOllamaCheck) {
await refreshOllamaSummary();
}
// 总是检查Ollama状态，因为这是独立于具体配置的
await refreshOllamaSummary();
// 检查已配置模型状态
await checkAllConfiguredModels();
@@ -2220,7 +2221,7 @@ const checkAllConfiguredModels = async () => {
if (formData.embedding.source === 'local' && formData.embedding.modelName && ollamaStatus.available) {
await checkAllOllamaModels();
} else if (formData.embedding.source === 'remote' && formData.embedding.modelName && formData.embedding.baseUrl) {
await checkRemoteModelStatus('embedding');
await checkEmbeddingModelStatus();
}
}
@@ -2243,6 +2244,51 @@ const onDimensionInput = (event: any) => {
formData.embedding.dimension = Number(event.target.value);
};
// 检查Embedding模型状态
const checkEmbeddingModelStatus = async () => {
if (!formData.embedding.modelName) {
return;
}
try {
modelStatus.embedding.checking = true;
modelStatus.embedding.checked = false;
modelStatus.embedding.available = false;
modelStatus.embedding.message = '';
const result = await testEmbeddingModel({
source: formData.embedding.source as 'local' | 'remote',
modelName: formData.embedding.modelName,
baseUrl: formData.embedding.source === 'remote' ? formData.embedding.baseUrl : undefined,
apiKey: formData.embedding.apiKey || undefined,
dimension: formData.embedding.dimension || undefined,
});
modelStatus.embedding.checked = true;
modelStatus.embedding.available = result.available || false;
modelStatus.embedding.message = result.message || '';
// 如果检测到维度信息，自动更新
if (result.available && result.dimension && result.dimension > 0) {
formData.embedding.dimension = result.dimension;
}
// 触发表单验证
setTimeout(() => {
form.value?.validate(['embedding.modelName']);
}, 100);
} catch (error) {
console.error('检查Embedding模型失败:', error);
modelStatus.embedding.checked = true;
modelStatus.embedding.available = false;
const err = error as any;
modelStatus.embedding.message = (err && err.message) || '检查失败';
} finally {
modelStatus.embedding.checking = false;
}
};
// 检测并自动填写 Embedding 维度
const detectEmbeddingDimension = async () => {
if (hasFiles.value) return;