init
This commit is contained in:
duanfuxiang
610
src/settings/AutoCompleteSettings.tsx
Normal file
610
src/settings/AutoCompleteSettings.tsx
Normal file
@@ -0,0 +1,610 @@
|
||||
import { Notice } from "obsidian";
|
||||
import * as React from "react";
|
||||
import { useState } from "react";
|
||||
|
||||
import {
|
||||
InfioSettings,
|
||||
} from '../types/settings';
|
||||
import { checkForErrors } from "../utils/auto-complete";
|
||||
|
||||
import CheckBoxSettingItem from "./components/CheckBoxSettingItem";
|
||||
import FewShotExampleSettings from "./components/FewShotExampleSettings";
|
||||
import SettingsItem from "./components/SettingsItem";
|
||||
import SliderSettingsItem from "./components/SliderSettingsItem";
|
||||
import TextSettingItem from "./components/TextSettingItem";
|
||||
import TriggerSettings from "./components/TriggerSettings";
|
||||
import {
|
||||
MAX_DELAY,
|
||||
MAX_FREQUENCY_PENALTY,
|
||||
MAX_MAX_CHAR_LIMIT,
|
||||
MAX_MAX_TOKENS,
|
||||
MAX_PRESENCE_PENALTY,
|
||||
MAX_TEMPERATURE,
|
||||
MAX_TOP_P,
|
||||
MIN_DELAY,
|
||||
MIN_FREQUENCY_PENALTY,
|
||||
MIN_MAX_CHAR_LIMIT,
|
||||
MIN_MAX_TOKENS,
|
||||
MIN_PRESENCE_PENALTY,
|
||||
MIN_TEMPERATURE,
|
||||
MIN_TOP_P
|
||||
} from "./versions";
|
||||
|
||||
interface IProps {
|
||||
onSettingsChanged(settings: InfioSettings): void;
|
||||
|
||||
settings: InfioSettings;
|
||||
}
|
||||
|
||||
export default function AutoCompleteSettings(props: IProps): React.JSX.Element {
|
||||
const [settings, _setSettings] = useState<InfioSettings>(props.settings);
|
||||
const errors = checkForErrors(settings);
|
||||
|
||||
React.useEffect(() => {
|
||||
_setSettings(props.settings);
|
||||
}, [props.settings]);
|
||||
|
||||
const updateSettings = (update: Partial<InfioSettings>) => {
|
||||
_setSettings((settings: InfioSettings) => {
|
||||
const newSettings = { ...settings, ...update };
|
||||
props.onSettingsChanged(newSettings);
|
||||
return newSettings;
|
||||
});
|
||||
};
|
||||
const resetSettings = () => {
|
||||
// const azureOAIApiSettings = settings.azureOAIApiSettings,
|
||||
// const openAIApiSettings = settings.openAIApiSettings,
|
||||
// const ollamaApiSettings = settings.ollamaApiSettings
|
||||
|
||||
// const newSettings: SmartCopilotSettings = {
|
||||
// ...settings
|
||||
// azureOAIApiSettings,
|
||||
// openAIApiSettings,
|
||||
// ollamaApiSettings,
|
||||
// advancedMode: settings.advancedMode,
|
||||
// };
|
||||
// updateSettings(newSettings);
|
||||
new Notice("Factory reset complete.");
|
||||
};
|
||||
|
||||
|
||||
// const renderAPISettings = () => {
|
||||
// if (settings.apiProvider === "azure") {
|
||||
// return (
|
||||
// <>
|
||||
// <TextSettingItem
|
||||
// name={"Azure OAI API URL"}
|
||||
// description={
|
||||
// "The Azure OpenAI services API URL is used in the requests."
|
||||
// }
|
||||
// placeholder={"Your API URL..."}
|
||||
// value={settings.azureOAIApiSettings.url}
|
||||
// errorMessage={errors.get("azureOAIApiSettings.url")}
|
||||
// setValue={(value: string) =>
|
||||
// updateSettings({
|
||||
// azureOAIApiSettings: {
|
||||
// ...settings.azureOAIApiSettings,
|
||||
// url: value,
|
||||
// },
|
||||
// })
|
||||
// }
|
||||
// />
|
||||
// <TextSettingItem
|
||||
// name={"Azure API key"}
|
||||
// description={
|
||||
// "The Azure OpenAI services API key used in the requests."
|
||||
// }
|
||||
// placeholder={"Your API key..."}
|
||||
// password
|
||||
// value={settings.azureOAIApiSettings.key}
|
||||
// errorMessage={errors.get("azureOAIApiSettings.key")}
|
||||
// setValue={(value: string) =>
|
||||
// updateSettings({
|
||||
// azureOAIApiSettings: {
|
||||
// ...settings.azureOAIApiSettings,
|
||||
// key: value,
|
||||
// },
|
||||
// })
|
||||
// }
|
||||
// />
|
||||
// <ConnectivityCheck key={"azure"} settings={settings} />
|
||||
// </>
|
||||
// );
|
||||
// }
|
||||
// if (settings.apiProvider === "openai") {
|
||||
// return (
|
||||
// <>
|
||||
// <TextSettingItem
|
||||
// name={"OpenAI API URL"}
|
||||
// description={
|
||||
// "The URL used in the requests."
|
||||
// }
|
||||
// placeholder={"Your API URL..."}
|
||||
// value={settings.openAIApiSettings.url}
|
||||
// errorMessage={errors.get("openAIApiSettings.url")}
|
||||
// setValue={(value: string) =>
|
||||
// updateSettings({
|
||||
// openAIApiSettings: {
|
||||
// ...settings.openAIApiSettings,
|
||||
// url: value,
|
||||
// },
|
||||
// })
|
||||
// }
|
||||
// />
|
||||
// <TextSettingItem
|
||||
// name={"OpenAI API key"}
|
||||
// description={"The API key used in the requests."}
|
||||
// placeholder={"Your API key..."}
|
||||
// password
|
||||
// value={settings.openAIApiSettings.key}
|
||||
// errorMessage={errors.get("openAIApiSettings.key")}
|
||||
// setValue={(value: string) =>
|
||||
// updateSettings({
|
||||
// openAIApiSettings: {
|
||||
// ...settings.openAIApiSettings,
|
||||
// key: value,
|
||||
// },
|
||||
// })
|
||||
// }
|
||||
// />
|
||||
// <TextSettingItem
|
||||
// name={"Model"}
|
||||
// description={"The value of the model parameter in the request body."}
|
||||
// placeholder="gpt-3.5-turbo"
|
||||
// value={settings.openAIApiSettings.model}
|
||||
// setValue={(value: string) =>
|
||||
// updateSettings({
|
||||
// openAIApiSettings: {
|
||||
// ...settings.openAIApiSettings,
|
||||
// model: value,
|
||||
// }
|
||||
// })
|
||||
// }
|
||||
// errorMessage={errors.get("openAIApiSettings.model")}
|
||||
// />
|
||||
|
||||
// <ConnectivityCheck key={"openai"} settings={settings} />
|
||||
// </>
|
||||
// );
|
||||
// }
|
||||
// if (settings.apiProvider === "ollama") {
|
||||
// return (
|
||||
// <>
|
||||
// <TextSettingItem
|
||||
// name={"API URL"}
|
||||
// description={
|
||||
// "The URL used in the requests."
|
||||
// }
|
||||
// placeholder={"Your API URL..."}
|
||||
// value={settings.ollamaApiSettings.url}
|
||||
// errorMessage={errors.get("ollamaApiSettings.url")}
|
||||
// setValue={(value: string) =>
|
||||
// updateSettings({
|
||||
// ollamaApiSettings: {
|
||||
// ...settings.ollamaApiSettings,
|
||||
// url: value,
|
||||
// },
|
||||
// })
|
||||
// }
|
||||
// />
|
||||
// <TextSettingItem
|
||||
// name={"Model"}
|
||||
// description={"The model you have locally running using OLLAMA."}
|
||||
// placeholder="Your model name..."
|
||||
// value={settings.ollamaApiSettings.model}
|
||||
// setValue={(value: string) =>
|
||||
// updateSettings({
|
||||
// ollamaApiSettings: {
|
||||
// ...settings.ollamaApiSettings,
|
||||
// model: value,
|
||||
// }
|
||||
// })
|
||||
// }
|
||||
// errorMessage={errors.get("ollamaApiSettings.model")}
|
||||
// />
|
||||
|
||||
// <ConnectivityCheck key={"openai"} settings={settings} />
|
||||
// </>
|
||||
// );
|
||||
// }
|
||||
// };
|
||||
|
||||
return (
|
||||
<div>
|
||||
<h2>AutoComplete</h2>
|
||||
<CheckBoxSettingItem
|
||||
name={"Enable"}
|
||||
description={
|
||||
"If disabled, nothing will trigger the extension or can result in an API call."
|
||||
}
|
||||
enabled={settings.autocompleteEnabled}
|
||||
setEnabled={(value) => updateSettings({ autocompleteEnabled: value })}
|
||||
/>
|
||||
<CheckBoxSettingItem
|
||||
name={"Cache completions"}
|
||||
description={
|
||||
"If disabled, the plugin will not cache the completions. After accepting or rejecting a completion, the plugin will not remember it. This might result in more API calls."
|
||||
}
|
||||
enabled={settings.cacheSuggestions}
|
||||
setEnabled={(value) => updateSettings({ cacheSuggestions: value })}
|
||||
/>
|
||||
{/* <DropDownSettingItem
|
||||
name={"API provider"}
|
||||
description={
|
||||
"The plugin supports multiple API providers. Each provider might require different settings."
|
||||
}
|
||||
value={settings.apiProvider}
|
||||
setValue={(value: string) => {
|
||||
if (value === "openai" || value === "azure" || value === "ollama") {
|
||||
updateSettings({ apiProvider: value });
|
||||
}
|
||||
}}
|
||||
options={{
|
||||
openai: "OpenAI API",
|
||||
azure: "Azure OAI API",
|
||||
ollama: "Self-hosted OLLAMA API"
|
||||
}}
|
||||
errorMessage={errors.get("apiProvider")}
|
||||
/> */}
|
||||
<CheckBoxSettingItem
|
||||
name={"Debug mode"}
|
||||
description={
|
||||
"If enabled, various debug messages will be logged to the console, such as the complete response from the API, including the chain of thought tokens."
|
||||
}
|
||||
enabled={settings.debugMode}
|
||||
setEnabled={(value) => updateSettings({ debugMode: value })}
|
||||
/>
|
||||
{/* <h2>API</h2>
|
||||
{renderAPISettings()} */}
|
||||
<h2>Model Options</h2>
|
||||
<SliderSettingsItem
|
||||
name={"Temperature"}
|
||||
description={
|
||||
"This parameter affects randomness in the sampling. Lower values result in more repetitive and deterministic responses. Higher temperatures will result in more unexpected or creative responses."
|
||||
}
|
||||
value={settings.modelOptions.temperature}
|
||||
errorMessage={errors.get("modelOptions.temperature")}
|
||||
setValue={(value: number) =>
|
||||
updateSettings({
|
||||
modelOptions: {
|
||||
...settings.modelOptions,
|
||||
temperature: value,
|
||||
},
|
||||
})
|
||||
}
|
||||
min={MIN_TEMPERATURE}
|
||||
max={MAX_TEMPERATURE}
|
||||
step={0.05}
|
||||
/>
|
||||
<SliderSettingsItem
|
||||
name={"TopP"}
|
||||
description={
|
||||
"Like the temperature parameter, the Top P parameter affects the randomness in sampling. Lowering the value will limit the model's token selection to likelier tokens while increasing the value expands the model's token selection with lower likelihood tokens."
|
||||
}
|
||||
value={settings.modelOptions.top_p}
|
||||
errorMessage={errors.get("modelOptions.top_p")}
|
||||
setValue={(value: number) =>
|
||||
updateSettings({
|
||||
modelOptions: {
|
||||
...settings.modelOptions,
|
||||
top_p: value,
|
||||
},
|
||||
})
|
||||
}
|
||||
min={MIN_TOP_P}
|
||||
max={MAX_TOP_P}
|
||||
step={0.05}
|
||||
/>
|
||||
{settings.apiProvider !== "ollama" && (<>
|
||||
<SliderSettingsItem
|
||||
name={"Frequency Penalty"}
|
||||
description={
|
||||
"This parameter reduces the chance of repeating a token proportionally based on how often it has appeared in the text so far. This decreases the likelihood of repeating the exact same text in a response."
|
||||
}
|
||||
value={settings.modelOptions.frequency_penalty}
|
||||
errorMessage={errors.get("modelOptions.frequency_penalty")}
|
||||
setValue={(value: number) =>
|
||||
updateSettings({
|
||||
modelOptions: {
|
||||
...settings.modelOptions,
|
||||
frequency_penalty: value,
|
||||
},
|
||||
})
|
||||
}
|
||||
min={MIN_FREQUENCY_PENALTY}
|
||||
max={MAX_FREQUENCY_PENALTY}
|
||||
step={0.05}
|
||||
/>
|
||||
<SliderSettingsItem
|
||||
name={"Presence Penalty"}
|
||||
description={
|
||||
"This parameter reduces the chance of repeating any token that has appeared in the text so far. This increases the likelihood of introducing new topics in a response."
|
||||
}
|
||||
value={settings.modelOptions.presence_penalty}
|
||||
errorMessage={errors.get("modelOptions.presence_penalty")}
|
||||
setValue={(value: number) =>
|
||||
updateSettings({
|
||||
modelOptions: {
|
||||
...settings.modelOptions,
|
||||
presence_penalty: value,
|
||||
},
|
||||
})
|
||||
}
|
||||
min={MIN_PRESENCE_PENALTY}
|
||||
max={MAX_PRESENCE_PENALTY}
|
||||
step={0.05}
|
||||
/>
|
||||
<SliderSettingsItem
|
||||
name={"Max Tokens"}
|
||||
description={
|
||||
"This parameter changes the maximum number of tokens the model is allowed to generate. This includes the chain of thought tokens before the answer."
|
||||
}
|
||||
value={settings.modelOptions.max_tokens}
|
||||
errorMessage={errors.get("modelOptions.max_tokens")}
|
||||
setValue={(value: number) =>
|
||||
updateSettings({
|
||||
modelOptions: {
|
||||
...settings.modelOptions,
|
||||
max_tokens: value,
|
||||
},
|
||||
})
|
||||
}
|
||||
min={MIN_MAX_TOKENS}
|
||||
max={MAX_MAX_TOKENS}
|
||||
step={10}
|
||||
/>
|
||||
</>)}
|
||||
|
||||
<h2>Preprocessing</h2>
|
||||
<CheckBoxSettingItem
|
||||
name={"Don't include dataviews"}
|
||||
description={
|
||||
"Dataview(js) blocks can be quite long while not providing much value to the AI. If this setting is enabled, data view blocks will be removed promptly to reduce the number of tokens. This could save you some money in the long run."
|
||||
}
|
||||
enabled={settings.dontIncludeDataviews}
|
||||
setEnabled={(value) =>
|
||||
updateSettings({ dontIncludeDataviews: value })
|
||||
}
|
||||
/>
|
||||
<SliderSettingsItem
|
||||
name={"Maximum Prefix Length"}
|
||||
description={
|
||||
"The maximum number of characters that will be included in the prefix. A larger value will increase the context for the completion, but it can also increase the cost or push you over the token limit."
|
||||
}
|
||||
value={settings.maxPrefixCharLimit}
|
||||
errorMessage={errors.get("maxPrefixCharLimit")}
|
||||
setValue={(value: number) =>
|
||||
updateSettings({ maxPrefixCharLimit: value })
|
||||
}
|
||||
min={MIN_MAX_CHAR_LIMIT}
|
||||
max={MAX_MAX_CHAR_LIMIT}
|
||||
step={100}
|
||||
suffix={" chars"}
|
||||
/>
|
||||
<SliderSettingsItem
|
||||
name={"Maximum Suffix Length"}
|
||||
description={
|
||||
"The maximum number of characters that will be included in the suffix. A larger value will increase the context for the completion, but it can also increase the cost or push you over the token limit."
|
||||
}
|
||||
value={settings.maxSuffixCharLimit}
|
||||
errorMessage={errors.get("maxSuffixCharLimit")}
|
||||
setValue={(value: number) =>
|
||||
updateSettings({ maxSuffixCharLimit: value })
|
||||
}
|
||||
min={MIN_MAX_CHAR_LIMIT}
|
||||
max={MAX_MAX_CHAR_LIMIT}
|
||||
step={100}
|
||||
suffix={" chars"}
|
||||
/>
|
||||
<h2>Postprocessing</h2>
|
||||
<CheckBoxSettingItem
|
||||
name={"Auto remove duplicate mat block indicators"}
|
||||
description={
|
||||
"The AI model might eagerly add a math block indicator ($), even though the cursor is already inside a math block. If this setting is enabled, the plugin will automatically remove these duplicate indicators from the completion."
|
||||
}
|
||||
enabled={settings.removeDuplicateMathBlockIndicator}
|
||||
setEnabled={(value) =>
|
||||
updateSettings({ removeDuplicateMathBlockIndicator: value })
|
||||
}
|
||||
/>
|
||||
<CheckBoxSettingItem
|
||||
name={"Auto remove duplicate mat block indicators"}
|
||||
description={
|
||||
"The AI model might eagerly add a code block indicator (`), even though the cursor is already inside a code block. If this setting is enabled, the plugin will automatically remove these duplicate indicators from the completion."
|
||||
}
|
||||
enabled={settings.removeDuplicateCodeBlockIndicator}
|
||||
setEnabled={(value) =>
|
||||
updateSettings({ removeDuplicateCodeBlockIndicator: value })
|
||||
}
|
||||
/>
|
||||
|
||||
<h2>Trigger</h2>
|
||||
<SliderSettingsItem
|
||||
name={"Delay"}
|
||||
description={
|
||||
"Delay in ms between the last character typed and the completion request."
|
||||
}
|
||||
value={settings.delay}
|
||||
errorMessage={errors.get("delay")}
|
||||
setValue={(value: number) => updateSettings({ delay: value })}
|
||||
min={MIN_DELAY}
|
||||
max={MAX_DELAY}
|
||||
step={100}
|
||||
suffix={"ms"}
|
||||
/>
|
||||
<TriggerSettings
|
||||
name={"Trigger words"}
|
||||
description={
|
||||
"Completions will be triggered if the text before the matches any of these words or characters. This can either be a direct string match or a regex match. When using a regex, make sure to include the end of line character ($)."
|
||||
}
|
||||
triggers={settings.triggers}
|
||||
setValues={(triggers) => updateSettings({ triggers })}
|
||||
errorMessage={errors.get("triggerWords")}
|
||||
errorMessages={errors}
|
||||
/>
|
||||
<h2>Privacy</h2>
|
||||
<SettingsItem
|
||||
name={"Ignored files"}
|
||||
description={
|
||||
<div>
|
||||
<p>This field enables you to specify files and directories that the plugin should ignore. When
|
||||
you open any of these files, the plugin will automatically disable itself and display a
|
||||
'disabled' status in the bottom menu. Enter one pattern per line. These patterns function
|
||||
similar to glob patterns. Here are some frequently used patterns:</p>
|
||||
<ul>
|
||||
<li><code>path/to/folder/**</code>: This pattern ignores all files and sub folders within
|
||||
this folder.
|
||||
</li>
|
||||
<li><code>"**/secret/**"</code>: This pattern ignores any file located inside a 'secret'
|
||||
directory,
|
||||
regardless of its location in the path.
|
||||
</li>
|
||||
<li><code>!path/to/folder/example.md</code>: This pattern explicitly undoes an ignore,
|
||||
making this file noticeable to the plugin.
|
||||
</li>
|
||||
<li><code>**/*Python*.md</code>: This pattern ignores any file with 'Python' in its name,
|
||||
irrespective of its location.
|
||||
</li>
|
||||
</ul>
|
||||
</div>
|
||||
}
|
||||
display={"block"}
|
||||
errorMessage={errors.get("ignoredFilePatterns")}
|
||||
>
|
||||
<textarea
|
||||
className="infio-autocomplete-setting-item-textarea"
|
||||
rows={10}
|
||||
placeholder="Your file patterns, e.g., **/secret/**"
|
||||
value={settings.ignoredFilePatterns}
|
||||
onChange={(e) =>
|
||||
updateSettings({
|
||||
ignoredFilePatterns: e.target.value
|
||||
})
|
||||
}
|
||||
/>
|
||||
</SettingsItem>
|
||||
<SettingsItem
|
||||
name={"Ignored tags"}
|
||||
description={
|
||||
<div>
|
||||
<p>Files containing any of these tags will be ignored. When you open a file containing a
|
||||
tag listed here, the plugin will automatically disable itself and display a 'disabled'
|
||||
status in the bottom menu. Enter one tag per line.
|
||||
</p>
|
||||
</div>
|
||||
}
|
||||
display={"block"}
|
||||
errorMessage={errors.get("ignoredTags")}
|
||||
>
|
||||
<textarea
|
||||
className="infio-autocomplete-setting-item-textarea"
|
||||
rows={10}
|
||||
placeholder="Your file tags, e.g., secret"
|
||||
value={settings.ignoredTags}
|
||||
onChange={(e) =>
|
||||
updateSettings({
|
||||
ignoredTags: e.target.value
|
||||
})
|
||||
}
|
||||
/>
|
||||
</SettingsItem>
|
||||
|
||||
<h2>Danger zone</h2>
|
||||
<SettingsItem
|
||||
name={"Factory Reset"}
|
||||
description={
|
||||
"Messed-up the settings? No worries, press this button! After that, the plugin will go back to the default settings. The URL and API key will remain unchanged."
|
||||
}
|
||||
>
|
||||
<button
|
||||
aria-label="Reset to default settings"
|
||||
onClick={resetSettings}
|
||||
>
|
||||
Reset
|
||||
</button>
|
||||
</SettingsItem>
|
||||
<CheckBoxSettingItem
|
||||
name={"Advanced mode"}
|
||||
description={
|
||||
"If you are familiar with prompt engineering, you can enable this setting to view the prompt generation and a few shot example settings. Turn off this button. It will not reset your changes; use the factory reset button for that."
|
||||
}
|
||||
enabled={settings.advancedMode}
|
||||
setEnabled={(value) => updateSettings({ advancedMode: value })}
|
||||
/>
|
||||
|
||||
{settings.advancedMode && (
|
||||
<>
|
||||
<h2>Advanced</h2>
|
||||
<TextSettingItem
|
||||
name={"Chain of thought removal regex"}
|
||||
description={
|
||||
"This regex is used to remove the chain of thought tokens from the generated answer. If it is not implemented correctly, the chain of thought tokens will be included in the suggested completion."
|
||||
}
|
||||
placeholder={"your regex..."}
|
||||
value={settings.chainOfThoughRemovalRegex}
|
||||
errorMessage={errors.get("chainOfThoughRemovalRegex")}
|
||||
setValue={(value: string) =>
|
||||
updateSettings({
|
||||
chainOfThoughRemovalRegex: value,
|
||||
})
|
||||
}
|
||||
/>
|
||||
|
||||
<SettingsItem
|
||||
name={"System Message"}
|
||||
description={
|
||||
"This system message gives the models all the context and instructions they need to complete the answer generation tasks. You can edit this message to your liking. If you edit the chain of thought formatting, make sure to update the extract regex and examples accordingly."
|
||||
}
|
||||
display={"block"}
|
||||
errorMessage={errors.get("systemMessage")}
|
||||
>
|
||||
<textarea
|
||||
className="infio-autocomplete-setting-item-textarea"
|
||||
rows={10}
|
||||
placeholder="Your system message..."
|
||||
value={settings.systemMessage}
|
||||
onChange={(e) =>
|
||||
updateSettings({
|
||||
systemMessage: e.target.value,
|
||||
})
|
||||
}
|
||||
/>
|
||||
</SettingsItem>
|
||||
|
||||
<SettingsItem
|
||||
name={"User Message template"}
|
||||
description={
|
||||
"This template defines how the prefix and suffix are formatted to create the user message. You have access to two variables: {{prefix}} and {{suffix}}. If you edit this, make sure to update the examples accordingly."
|
||||
}
|
||||
display={"block"}
|
||||
errorMessage={errors.get("userMessageTemplate")}
|
||||
>
|
||||
<textarea
|
||||
className="infio-autocomplete-setting-item-textarea"
|
||||
rows={3}
|
||||
placeholder="{{prefix}}<mask/>{{suffix}}"
|
||||
value={settings.userMessageTemplate}
|
||||
onChange={(e) =>
|
||||
updateSettings({
|
||||
userMessageTemplate: e.target.value,
|
||||
})
|
||||
}
|
||||
/>
|
||||
</SettingsItem>
|
||||
<FewShotExampleSettings
|
||||
fewShotExamples={settings.fewShotExamples}
|
||||
name={"Few Shot Examples"}
|
||||
description={
|
||||
"The model uses these examples to learn the expected answer format. Not all examples are sent at the same time. We only send the relevant examples, given the current cursor location. For example, the CodeBlock examples are only sent if the cursor is in a code block. If no special context is detected, we send the Text examples. Each context has a default of 2 examples, but you can add or remove examples if there is at least one per context. You can add more examples, but this will increase the inference costs."
|
||||
}
|
||||
setFewShotExamples={(value) =>
|
||||
updateSettings({ fewShotExamples: value })
|
||||
}
|
||||
errorMessages={errors}
|
||||
|
||||
/>
|
||||
</>
|
||||
)}
|
||||
</div>
|
||||
);
|
||||
}
|
||||
Reference in New Issue
Block a user