Skip to content

fix: default whisper model to QuantizedLargeV3Turbo #706

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
data-bot-coasys wants to merge 8 commits into
base: dev
Choose a base branch
from
+49 −17
Open

fix: default whisper model to QuantizedLargeV3Turbo #706

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
227262a
fix: default whisper model to QuantizedLargeV3Turbo
data-bot-coasys Mar 3, 2026
c68f4da
Merge branch 'dev' into fix/whisper-default-model
lucksus Mar 3, 2026
9ae20c5
fix: show whisper tiny quantized in all launcher setup modes
data-bot-coasys Mar 3, 2026
5bee0b5
fix: use WHISPER_MODEL constant as fallback in get_whisper_model_size
data-bot-coasys Mar 4, 2026
9c43095
fix: revert get_whisper_model_size fallback to Tiny for CI compatibility
data-bot-coasys Mar 4, 2026
977652b
fix: use WHISPER_MODEL constant in fallback path and deduplicate UI l…
data-bot-coasys Mar 5, 2026
d5d328b
fix: remove redundant whisperModel assignment in Local mode
data-bot-coasys Mar 6, 2026
83fe3b6
fix: set default TRANSCRIPTION model after adding Whisper model
data-bot-coasys Mar 12, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions rust-executor/src/ai_service/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,7 +32,7 @@ use log::error;

pub type Result<T> = std::result::Result<T, AnyError>;

static WHISPER_MODEL: WhisperSource = WhisperSource::Small;
static WHISPER_MODEL: WhisperSource = WhisperSource::QuantizedLargeV3Turbo;
static TRANSCRIPTION_TIMEOUT_SECS: u64 = 30; // 30 seconds (was 2 minutes)
static TRANSCRIPTION_CHECK_INTERVAL_SECS: u64 = 5; // 5 seconds (was 10)

Expand Down Expand Up @@ -1105,8 +1105,8 @@ impl AIService {
}
}

// Default to tiny if nothing found
Ok(WhisperSource::Tiny)
// Default to the configured WHISPER_MODEL constant when no model is persisted in the DB.
Ok(WHISPER_MODEL)
}

pub async fn open_transcription_stream(
Expand Down
60 changes: 46 additions & 14 deletions ui/src/components/Login.tsx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,10 @@ import { open } from "@tauri-apps/plugin-shell";
import "../index.css";
import Logo from "./Logo";

const WHISPER_TURBO_URL =
"https://huggingface.co/Demonthos/candle-quantized-whisper-large-v3-turbo";
const WHISPER_TURBO_LABEL = "Whisper large v3 turbo quantized (454MB)";

const Login = () => {
const {
state: { loading, hasLoginError },
Expand Down Expand Up @@ -165,12 +169,11 @@ const Login = () => {
}

async function saveModels() {
let whisperModel = "whisper_small";
let whisperModel = "whisper_large_v3_turbo_quantized";
// add llm model
if (aiMode !== "None") {
const llm = { modelType: "LLM" } as ModelInput;
if (aiMode === "Local") {
whisperModel = "whisper_distil_large_v3";
llm.name = "Qwen2.5.1-Coder-7B-Instruct";
llm.local = { fileName: "Qwen2.5.1-Coder-7B-Instruct" };
} else {
Expand All @@ -192,12 +195,14 @@ const Login = () => {
local: { fileName: "bert" },
modelType: "EMBEDDING",
});
// add medium whisper model
client!.ai.addModel({
name: "Whisper",
local: { fileName: whisperModel },
modelType: "TRANSCRIPTION",
});
// add main whisper model (quantized large v3 turbo)
client!.ai
.addModel({
name: "Whisper",
local: { fileName: whisperModel },
modelType: "TRANSCRIPTION",
})
.then((modelId) => client!.ai.setDefaultModel("TRANSCRIPTION", modelId));

client!.ai.addModel({
name: "Whisper tiny quantized",
Expand Down Expand Up @@ -640,15 +645,24 @@ const Login = () => {
</a>
</p>
and
<p>
<a
onClick={() =>
open(WHISPER_TURBO_URL)
}
style={{ cursor: "pointer" }}
>{WHISPER_TURBO_LABEL}</a>
</p>
and
<p>
<a
onClick={() =>
open(
"https://huggingface.co/openai/whisper-large-v3-turbo"
"https://huggingface.co/lmz/candle-whisper"
)
}
style={{ cursor: "pointer" }}
>Whisper distill large v3 (1.42GB)</a>
>Whisper tiny quantized (42MB)</a>
</p>
and
<p>
Expand Down Expand Up @@ -814,15 +828,24 @@ const Login = () => {

<j-text>
This will still download
<p>
<a
onClick={() =>
open(WHISPER_TURBO_URL)
}
style={{ cursor: "pointer" }}
>{WHISPER_TURBO_LABEL}</a>
</p>
and
<p>
<a
onClick={() =>
open(
"https://huggingface.co/openai/whisper-small"
"https://huggingface.co/lmz/candle-whisper"
)
}
style={{ cursor: "pointer" }}
>Whisper small (244MB)</a>
>Whisper tiny quantized (42MB)</a>
</p>
and
<p>
Expand Down Expand Up @@ -856,15 +879,24 @@ const Login = () => {

<j-text>
This will still download
<p>
<a
onClick={() =>
open(WHISPER_TURBO_URL)
}
style={{ cursor: "pointer" }}
>{WHISPER_TURBO_LABEL}</a>
</p>
and
<p>
<a
onClick={() =>
open(
"https://huggingface.co/openai/whisper-small"
"https://huggingface.co/lmz/candle-whisper"
)
}
style={{ cursor: "pointer" }}
>Whisper small (244MB)</a>
>Whisper tiny quantized (42MB)</a>
</p>
and
<p>
Expand Down