mirror of
https://github.com/huggingface/text-generation-inference.git
synced 2025-09-11 12:24:53 +00:00
feat: support do_sample param in ChatRequest
This commit is contained in:
parent
06c3d4b1ec
commit
0520bde039
@ -709,6 +709,10 @@ pub(crate) struct ChatRequest {
|
|||||||
#[schema(nullable = true, example = "null")]
|
#[schema(nullable = true, example = "null")]
|
||||||
#[serde(deserialize_with = "deserialize_tool_choice::deserialize")]
|
#[serde(deserialize_with = "deserialize_tool_choice::deserialize")]
|
||||||
pub tool_choice: Option<ToolType>,
|
pub tool_choice: Option<ToolType>,
|
||||||
|
|
||||||
|
#[serde(default)]
|
||||||
|
#[schema(default = "false", example = true)]
|
||||||
|
pub do_sample: bool,
|
||||||
}
|
}
|
||||||
|
|
||||||
fn default_tool_prompt() -> Option<String> {
|
fn default_tool_prompt() -> Option<String> {
|
||||||
|
@ -1060,7 +1060,7 @@ async fn chat_completions(
|
|||||||
top_k: None,
|
top_k: None,
|
||||||
top_p: req.top_p,
|
top_p: req.top_p,
|
||||||
typical_p: None,
|
typical_p: None,
|
||||||
do_sample: true,
|
do_sample: req.do_sample,
|
||||||
max_new_tokens,
|
max_new_tokens,
|
||||||
return_full_text: None,
|
return_full_text: None,
|
||||||
stop,
|
stop,
|
||||||
|
Loading…
Reference in New Issue
Block a user