Skip to content

fix: change kwargs for akera ASR model #40

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

Merged
merged 4 commits into from
Sep 3, 2025
+36 −3
Merged

fix: change kwargs for akera ASR model #40

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
e2c487b
fix: change kwargs for akera ASR model
nikochiko Aug 28, 2025
8439f4d
feat: add akera to deployment chart
nikochiko Sep 2, 2025
911468b
revert: keep old env vars for model ids in dev script
nikochiko Sep 2, 2025
68fe01e
fix: update limits value for long-queue deployment for akera
nikochiko Sep 2, 2025
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
7 changes: 4 additions & 3 deletions api.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -99,10 +99,11 @@ class ImageCaptioningInput(BaseModel):

class WhisperInputs(BaseModel):
audio: str
task: typing.Literal["translate", "transcribe"] = "transcribe"
language: str = None
task: typing.Literal["translate", "transcribe"] | None = None
language: str | None = None
return_timestamps: bool = False
decoder_kwargs: dict = None
max_length: int | None = None
decoder_kwargs: dict | None = None

chunk_length_s: float = 30
stride_length_s: typing.Tuple[float, float] = (6, 0)
Expand Down
2 changes: 2 additions & 0 deletions common/whisper.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,8 @@ def whisper(pipeline: PipelineInfo, inputs: WhisperInputs) -> AsrOutput:
generate_kwargs["language"] = inputs.language
if inputs.task:
generate_kwargs["task"] = inputs.task
if inputs.max_length:
generate_kwargs["max_length"] = inputs.max_length
if generate_kwargs:
kwargs["generate_kwargs"] = generate_kwargs

Expand Down
61 changes: 3 additions & 58 deletions scripts/run-dev.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,64 +19,9 @@ docker build . -f $VARIANT/Dockerfile -t $IMG
docker rm -f $IMG || true
docker run \
-e IMPORTS=$IMPORTS \
-e WAV2LIP_MODEL_IDS="
Copy link
Member

@devxpy devxpy Sep 2, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

keep these around? easy to test

Copy link
Member Author

@nikochiko nikochiko Sep 2, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yes, will revert this one. i did have to add WIHSPER_TOKENIZER_FROM for this model to work. we don't add that for any other whisper model

wav2lip_gan.pth
" \
-e DEFORUM_MODEL_IDS="
Protogen_V2.2.ckpt
"\
-e EMBEDDING_MODEL_IDS="
intfloat/e5-large-v2
intfloat/e5-base-v2
intfloat/multilingual-e5-base
"\
-e MMS_MODEL_IDS="
facebook/mms-1b-all
"\
-e WHISPER_MODEL_IDS="
dmatekenya/whisper-large-v3-chichewa
" \
-e WHISPER_TOKENIZER_FROM="
openai/whisper-large-v3
"\
-e SD_MODEL_IDS="
stabilityai/stable-diffusion-2-inpainting
runwayml/stable-diffusion-inpainting
" \
-e CONTROLNET_MODEL_IDS="
lllyasviel/sd-controlnet-canny
lllyasviel/sd-controlnet-depth
lllyasviel/sd-controlnet-hed
lllyasviel/sd-controlnet-mlsd
lllyasviel/sd-controlnet-normal
lllyasviel/sd-controlnet-openpose
lllyasviel/sd-controlnet-scribble
lllyasviel/sd-controlnet-seg
lllyasviel/control_v11f1e_sd15_tile
ioclab/control_v1p_sd15_brightness
monster-labs/control_v1p_sd15_qrcode_monster/v2
" \
-e DIS_MODEL_IDS="
isnet-general-use.pth
"\
-e U2NET_MODEL_IDS="
u2net
"\
-e SEAMLESS_MODEL_IDS="
facebook/seamless-m4t-v2-large
"\
-e SADTALKER_MODEL_IDS="
SadTalker_V0.0.2_512.safetensors
"\
-e GFPGAN_MODEL_IDS="
GFPGANv1.4
"\
-e ESRGAN_MODEL_IDS="
RealESRGAN_x2plus
"\
-e LLM_MODEL_IDS="
aisingapore/llama3-8b-cpt-sea-lionv2-instruct
"\
-e QUEUE_PREFIX="gooey-gpu/short" \
-e WHISPER_MODEL_IDS=akera/whisper-large-v3-kik-full_v2 \
-e WHISPER_TOKENIZER_FROM=akera/whisper-large-v3-kik-full_v2 \
-e C_FORCE_ROOT=1 \
-e BROKER_URL=${BROKER_URL:-"amqp://"} \
-e RESULT_BACKEND=${RESULT_BACKEND:-"redis://"} \
Expand Down
Loading