Rm jsonMode
Browse files
src/lib/components/InferencePlayground/InferencePlayground.svelte
CHANGED
@@ -21,7 +21,7 @@
|
|
21 |
{
|
22 |
id: String(Math.random()),
|
23 |
model: '01-ai/Yi-1.5-34B-Chat',
|
24 |
-
config: { temperature: 0.5, maxTokens: 2048, streaming: true
|
25 |
messages: startMessages
|
26 |
}
|
27 |
];
|
@@ -128,7 +128,6 @@
|
|
128 |
requestMessages,
|
129 |
conversation.config.temperature,
|
130 |
conversation.config.maxTokens,
|
131 |
-
conversation.config.jsonMode,
|
132 |
(content) => {
|
133 |
if (streamingMessage) {
|
134 |
streamingMessage.content = content;
|
@@ -146,7 +145,6 @@
|
|
146 |
requestMessages,
|
147 |
conversation.config.temperature,
|
148 |
conversation.config.maxTokens,
|
149 |
-
conversation.config.jsonMode
|
150 |
);
|
151 |
// check if the user did not abort the request
|
152 |
if (waitForNonStreaming) {
|
@@ -386,7 +384,7 @@
|
|
386 |
{
|
387 |
id: String(Math.random()),
|
388 |
model: e.target.value,
|
389 |
-
config: { temperature: 0.5, maxTokens: 2048, streaming: true
|
390 |
messages: [...conversations[0].messages]
|
391 |
}
|
392 |
];
|
|
|
21 |
{
|
22 |
id: String(Math.random()),
|
23 |
model: '01-ai/Yi-1.5-34B-Chat',
|
24 |
+
config: { temperature: 0.5, maxTokens: 2048, streaming: true },
|
25 |
messages: startMessages
|
26 |
}
|
27 |
];
|
|
|
128 |
requestMessages,
|
129 |
conversation.config.temperature,
|
130 |
conversation.config.maxTokens,
|
|
|
131 |
(content) => {
|
132 |
if (streamingMessage) {
|
133 |
streamingMessage.content = content;
|
|
|
145 |
requestMessages,
|
146 |
conversation.config.temperature,
|
147 |
conversation.config.maxTokens,
|
|
|
148 |
);
|
149 |
// check if the user did not abort the request
|
150 |
if (waitForNonStreaming) {
|
|
|
384 |
{
|
385 |
id: String(Math.random()),
|
386 |
model: e.target.value,
|
387 |
+
config: { temperature: 0.5, maxTokens: 2048, streaming: true },
|
388 |
messages: [...conversations[0].messages]
|
389 |
}
|
390 |
];
|
src/lib/components/InferencePlayground/InferencePlaygroundGenerationConfig.svelte
CHANGED
@@ -1,9 +1,4 @@
|
|
1 |
<script lang="ts">
|
2 |
-
// export let temperature = 0.5;
|
3 |
-
// export let maxTokens = 2048;
|
4 |
-
// export let streaming = true;
|
5 |
-
// export let jsonMode = true;
|
6 |
-
|
7 |
export let config;
|
8 |
export let classNames = '';
|
9 |
</script>
|
|
|
1 |
<script lang="ts">
|
|
|
|
|
|
|
|
|
|
|
2 |
export let config;
|
3 |
export let classNames = '';
|
4 |
</script>
|
src/lib/components/InferencePlayground/inferencePlaygroundUtils.ts
CHANGED
@@ -18,7 +18,6 @@ export async function handleStreamingResponse(
|
|
18 |
messages: ChatCompletionInputMessage[],
|
19 |
temperature: number,
|
20 |
maxTokens: number,
|
21 |
-
jsonMode: boolean,
|
22 |
onChunk: (content: string) => void,
|
23 |
abortController: AbortController
|
24 |
): Promise<void> {
|
@@ -30,7 +29,6 @@ export async function handleStreamingResponse(
|
|
30 |
messages: messages,
|
31 |
temperature: temperature,
|
32 |
max_tokens: maxTokens,
|
33 |
-
json_mode: jsonMode
|
34 |
},
|
35 |
{ signal: abortController.signal }
|
36 |
)) {
|
@@ -54,14 +52,12 @@ export async function handleNonStreamingResponse(
|
|
54 |
messages: ChatCompletionInputMessage[],
|
55 |
temperature: number,
|
56 |
maxTokens: number,
|
57 |
-
jsonMode: boolean
|
58 |
): Promise<ChatCompletionInputMessage> {
|
59 |
const response = await hf.chatCompletion({
|
60 |
model: model,
|
61 |
messages: messages,
|
62 |
temperature: temperature,
|
63 |
max_tokens: maxTokens,
|
64 |
-
json_mode: jsonMode
|
65 |
});
|
66 |
|
67 |
if (response.choices && response.choices.length > 0) {
|
|
|
18 |
messages: ChatCompletionInputMessage[],
|
19 |
temperature: number,
|
20 |
maxTokens: number,
|
|
|
21 |
onChunk: (content: string) => void,
|
22 |
abortController: AbortController
|
23 |
): Promise<void> {
|
|
|
29 |
messages: messages,
|
30 |
temperature: temperature,
|
31 |
max_tokens: maxTokens,
|
|
|
32 |
},
|
33 |
{ signal: abortController.signal }
|
34 |
)) {
|
|
|
52 |
messages: ChatCompletionInputMessage[],
|
53 |
temperature: number,
|
54 |
maxTokens: number,
|
|
|
55 |
): Promise<ChatCompletionInputMessage> {
|
56 |
const response = await hf.chatCompletion({
|
57 |
model: model,
|
58 |
messages: messages,
|
59 |
temperature: temperature,
|
60 |
max_tokens: maxTokens,
|
|
|
61 |
});
|
62 |
|
63 |
if (response.choices && response.choices.length > 0) {
|
src/lib/types/index.d.ts
CHANGED
@@ -7,7 +7,6 @@ type GenerationConfig = {
|
|
7 |
temperature: number;
|
8 |
maxTokens: number;
|
9 |
streaming: boolean;
|
10 |
-
jsonMode: boolean;
|
11 |
};
|
12 |
|
13 |
type Conversation = {
|
|
|
7 |
temperature: number;
|
8 |
maxTokens: number;
|
9 |
streaming: boolean;
|
|
|
10 |
};
|
11 |
|
12 |
type Conversation = {
|