[whisper] allow to pass text/audio specific kwargs

[eustlb]

What does this PR do?

This allows to do something like:

from transformers import AutoProcessor
from datasets import Audio, load_dataset

model_id = "openai/whisper-large-v3"
processor = AutoProcessor.from_pretrained(model_id)

dataset = load_dataset("hf-internal-testing/librispeech_asr_dummy", "clean", split="validation")
dataset = dataset.cast_column("audio", Audio(processor.feature_extractor.sampling_rate))
sample = dataset[0]["audio"]

inputs = processor(
    audio=sample["array"],
    text="this is the transcription text",
    audio_kwargs={
        "max_length": 10000,
    },
    text_kwargs={
        "max_length": 5,
    },
    common_kwargs={
        "return_tensors": "pt",
    }
)
for k, v in inputs.items():
    print(f"{k}: {v.shape}")

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[ArthurZucker]

Okayyy not sure a test is necessary?

[eustlb]

run-slow: whisper

[github-actions]

[For maintainers] Suggested jobs to run (before merge)

run-slow: whisper

[github-actions]

This comment contains run-slow, running the specified jobs:

models: ["models/whisper"]
quantizations: []

[github-actions]

CI Results

Workflow Run ⚙️

Commit Info

Context	Commit	Description
RUN	be16e231	merge commit
PR	4d187bfc	branch commit
main	b7b9d252	base commit

Model CI Report

❌ 1 new failed tests from this PR 😭

• whisper:
  tests/models/whisper/test_processing_whisper.py::WhisperProcessorTest::test_tokenizer