Dedicated Endpoints

Run this model inference on single tenant GPU with unmatched speed and reliability at scale.

Learn more
Container

Run this model inference with full control and performance in your environment.

Learn more

Get help setting up a custom Dedicated Endpoints.

Talk with our engineer to get a quote for reserved GPU instances with discounts.

README

License: apache-2.0

Training eval (imaginary-jail 30m holdout, 235 segments)

MetricValue
WER9.96%
CER3.31%

Benchmark results (normalized ASCII IPA)

ILSpeech test (data/ilspeech-v2/test, 150 samples)

MetricBase (whisper-he-ipa)This model
CER2.17%1.73%
WER9.55%7.99%
SER8.22%6.63%
VER1.97%1.69%
Exact match44.7%46.7%

Michael Gold v1 (data/michael-gold-v1, 561 samples)

MetricBase (whisper-he-ipa)This model
CER4.20%4.05%
WER19.71%19.12%
SER14.55%14.98%
VER2.79%2.80%
Exact match6.6%7.7%

Full per-sample reports are in benchmarks/.

Usage

python

from transformers import pipeline
pipe = pipeline(
"automatic-speech-recognition",
model="aunikud/whisper-heb-ipa",
generate_kwargs={"language": "he", "task": "transcribe"},
)
print(pipe("audio.wav")["text"])

CLI from this repo:

bash

uv run src/infer.py audio.wav --model aunikud/whisper-heb-ipa

Training data

  • data/ilspeech-v2/train
  • data/imaginary-jail-clean-v2 train split (metadata_train.csv, ~11h)

Eval during training used metadata_eval_30m.csv (30-minute holdout from the same dataset).

Model provider

aunikud

Model tree

Base

aunikud/whisper-he-ipa

Fine-tuned

this model

Modalities

Input

Audio

Output

Text

Pricing

Dedicated Endpoints

View details

Supported Functionality

Model APIs

Dedicated Endpoints

Container

More information

Explore FriendliAI today