server: sampling overrides on the speech endpoint
The speech body accepts seed, max_new_tokens, temperature, top_k, top_p, and repetition_penalty. Unset fields keep the engine defaults, a temperature of zero selects greedy decoding, and the subtalker mirrors the talker knobs. A fixed seed makes a request reproducible.
This commit is contained in:
@@ -626,7 +626,13 @@ Endpoints :
|
||||
```
|
||||
POST /v1/audio/speech OAI text-to-speech; response_format "pcm"
|
||||
streams s16le 24 kHz mono chunked as it is
|
||||
generated, "wav" returns a one-shot RIFF file
|
||||
generated, "wav" returns a one-shot RIFF file.
|
||||
Optional sampling overrides ride in the same
|
||||
body: seed, max_new_tokens, temperature,
|
||||
top_k, top_p, repetition_penalty. Unset
|
||||
fields keep the engine defaults, temperature
|
||||
0 selects greedy decoding, the subtalker
|
||||
mirrors the talker knobs
|
||||
GET /v1/models single loaded model
|
||||
GET /v1/voices model speakers plus registered cloned voices
|
||||
POST /v1/voices register a cloned voice: {name, ref_text,
|
||||
|
||||
Reference in New Issue
Block a user