rixty/SillyTavern
1
0
Fork 0
mirror of https://github.com/SillyTavern/SillyTavern.git synced 2025-06-05 21:59:27 +02:00
Code Issues Packages Projects Releases Wiki Activity

(beta) Add davinci models to OpenAI

Browse Source
This commit is contained in:
Cohee
2023-07-06 13:18:33 +03:00
parent 35390c07ef
commit 85e82f854b
3 changed files with 67 additions and 38 deletions
Download Patch File Download Diff File Expand all files Collapse all files

51
public/scripts/openai.js
View File

@@ -81,11 +81,11 @@ const default_bias_presets = {
]
};
const gpt3_max = 4095;
const gpt3_16k_max = 16383;
const gpt4_max = 8191;
const gpt_neox_max = 2048;
const gpt4_32k_max = 32767;
const max_2k = 2047;
const max_4k = 4095;
const max_8k = 8191;
const max_16k = 16383;
const max_32k = 32767;
const scale_max = 7900; // Probably more. Save some for the system prompt defined on Scale site.
const claude_max = 8000; // We have a proper tokenizer, so theoretically could be larger (up to 9k)
const palm2_max = 7500; // The real context window is 8192, spare some for padding due to using turbo tokenizer
@@ -112,7 +112,7 @@ const default_settings = {
top_p_openai: 1.0,
top_k_openai: 0,
stream_openai: false,
openai_max_context: gpt3_max,
openai_max_context: max_4k,
openai_max_tokens: 300,
nsfw_toggle: true,
enhance_definitions: false,
@@ -147,7 +147,7 @@ const oai_settings = {
top_p_openai: 1.0,
top_k_openai: 0,
stream_openai: false,
openai_max_context: gpt3_max,
openai_max_context: max_4k,
openai_max_tokens: 300,
nsfw_toggle: true,
enhance_definitions: false,
@@ -689,6 +689,7 @@ async function sendOpenAIRequest(type, openai_msgs_tosend, signal) {
const isClaude = oai_settings.chat_completion_source == chat_completion_sources.CLAUDE;
const isOpenRouter = oai_settings.use_openrouter && oai_settings.chat_completion_source == chat_completion_sources.WINDOWAI;
const isScale = oai_settings.chat_completion_source == chat_completion_sources.SCALE;
const isTextCompletion = oai_settings.chat_completion_source == chat_completion_sources.OPENAI && (oai_settings.openai_model.startsWith('text-') || oai_settings.openai_model.startsWith('code-'));
const stream = type !== 'quiet' && oai_settings.stream_openai && !isScale;
// If we're using the window.ai extension, use that instead
@@ -804,7 +805,7 @@ async function sendOpenAIRequest(type, openai_msgs_tosend, signal) {
throw new Error(data);
}
return data.choices[0]["message"]["content"];
return !isTextCompletion ? data.choices[0]["message"]["content"] : data.choices[0]["text"];
}
}
@@ -812,7 +813,7 @@ function getStreamingReply(getMessage, data) {
if (oai_settings.chat_completion_source == chat_completion_sources.CLAUDE) {
getMessage = data.completion || "";
} else {
getMessage += data.choices[0]?.delta?.content || data.choices[0]?.message?.content || "";
getMessage += data.choices[0]?.delta?.content || data.choices[0]?.message?.content || data.choices[0]?.text || "";
}
return getMessage;
}
@@ -1670,26 +1671,26 @@ async function onModelChange() {
$('#openai_max_context').attr('max', claude_max);
}
else if (value.includes('gpt-3.5-turbo-16k')) {
$('#openai_max_context').attr('max', gpt3_16k_max);
$('#openai_max_context').attr('max', max_16k);
}
else if (value.includes('gpt-3.5')) {
$('#openai_max_context').attr('max', gpt3_max);
$('#openai_max_context').attr('max', max_4k);
}
else if (value.includes('gpt-4-32k')) {
$('#openai_max_context').attr('max', gpt4_32k_max);
$('#openai_max_context').attr('max', max_32k);
}
else if (value.includes('gpt-4')) {
$('#openai_max_context').attr('max', gpt4_max);
$('#openai_max_context').attr('max', max_8k);
}
else if (value.includes('palm-2')) {
$('#openai_max_context').attr('max', palm2_max);
}
else if (value.includes('GPT-NeoXT')) {
$('#openai_max_context').attr('max', gpt_neox_max);
$('#openai_max_context').attr('max', max_2k);
}
else {
// default to gpt-3 (4095 tokens)
$('#openai_max_context').attr('max', gpt3_max);
$('#openai_max_context').attr('max', max_4k);
}
oai_settings.openai_max_context = Math.min(Number($('#openai_max_context').attr('max')), oai_settings.openai_max_context);
@@ -1709,17 +1710,23 @@ async function onModelChange() {
if (oai_settings.max_context_unlocked) {
$('#openai_max_context').attr('max', unlocked_max);
}
else if (value == 'gpt-4' || value == 'gpt-4-0314' || value == 'gpt-4-0613') {
$('#openai_max_context').attr('max', gpt4_max);
else if (['gpt-4', 'gpt-4-0314', 'gpt-4-0613'].includes(value)) {
$('#openai_max_context').attr('max', max_8k);
}
else if (value == 'gpt-4-32k' || value == 'gpt-4-32k-0314' || value == 'gpt-4-32k-0613') {
$('#openai_max_context').attr('max', gpt4_32k_max);
else if (['gpt-4-32k', 'gpt-4-32k-0314', 'gpt-4-32k-0613'].includes(value)) {
$('#openai_max_context').attr('max', max_32k);
}
else if (value == 'gpt-3.5-turbo-16k' || value == 'gpt-3.5-turbo-16k-0613') {
$('#openai_max_context').attr('max', gpt3_16k_max);
else if (['gpt-3.5-turbo-16k', 'gpt-3.5-turbo-16k-0613'].includes(value)) {
$('#openai_max_context').attr('max', max_16k);
}
else if (value == 'code-davinci-002') {
$('#openai_max_context').attr('max', max_8k);
}
else if (['text-curie-001', 'text-babbage-001', 'text-ada-001'].includes(value)) {
$('#openai_max_context').attr('max', max_2k);
}
else {
$('#openai_max_context').attr('max', gpt3_max);
$('#openai_max_context').attr('max', max_4k);
}
oai_settings.openai_max_context = Math.min(oai_settings.openai_max_context, Number($('#openai_max_context').attr('max')));