Datasets:

argmaxinc
/

whisperkit-0.7.0-evals

+---
+pretty_name: "WhisperKit ASR Evaluation Results"
+viewer: false
+library_name: whisperkit
+tags:
+- whisper
+- whisperkit
+- coreml
+- asr
+- quantized
+---
+# WhisperKit Transcription Quality
+## Dataset: `earnings22-12hours`
+abc
+|                                                                                                       | WER (↓)                                                                                                                                   |   QoI (↑) |   File Size (MB) | Code Commit                                                    |
+|:------------------------------------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------|----------:|-----------------:|:---------------------------------------------------------------|
+| [large-v3_turbo](https://hf.co/argmaxinc/whisperkit-coreml/tree/main/openai_whisper-large-v3_turbo)   | [11.97](https://hf.co/datasets/argmaxinc/whisperkit-evals-staging/tree/main/WhisperKit/openai_whisper-large-v3_turbo/earnings22-12hours)  |     100   |             3100 | [Link](https://github.com/argmaxinc/WhisperKit/commit/65cb888) |
+| [large-v2](https://hf.co/argmaxinc/whisperkit-coreml/tree/main/openai_whisper-large-v2)               | [12.4](https://hf.co/datasets/argmaxinc/whisperkit-evals-staging/tree/main/WhisperKit/openai_whisper-large-v2/earnings22-12hours)         |      38.5 |             3100 | [Link](https://github.com/argmaxinc/WhisperKit/commit/65cb888) |
+| [large-v2_turbo](https://hf.co/argmaxinc/whisperkit-coreml/tree/main/openai_whisper-large-v2_turbo)   | [12.47](https://hf.co/datasets/argmaxinc/whisperkit-evals-staging/tree/main/WhisperKit/openai_whisper-large-v2_turbo/earnings22-12hours)  |      38.5 |             3100 | [Link](https://github.com/argmaxinc/WhisperKit/commit/65cb888) |
+| [distil-large-v3](https://hf.co/argmaxinc/whisperkit-coreml/tree/main/distil-whisper_distil-large-v3) | [12.32](https://hf.co/datasets/argmaxinc/whisperkit-evals-staging/tree/main/WhisperKit/distil-whisper_distil-large-v3/earnings22-12hours) |      23.1 |             1510 | [Link](https://github.com/argmaxinc/WhisperKit/commit/65cb888) |
+| [small.en](https://hf.co/argmaxinc/whisperkit-coreml/tree/main/openai_whisper-small.en)               | [13.08](https://hf.co/datasets/argmaxinc/whisperkit-evals-staging/tree/main/WhisperKit/openai_whisper-small.en/earnings22-12hours)        |      15.4 |              483 | [Link](https://github.com/argmaxinc/WhisperKit/commit/65cb888) |
+| [small](https://hf.co/argmaxinc/whisperkit-coreml/tree/main/openai_whisper-small)                     | [13.27](https://hf.co/datasets/argmaxinc/whisperkit-evals-staging/tree/main/WhisperKit/openai_whisper-small/earnings22-12hours)           |      15.4 |              483 | [Link](https://github.com/argmaxinc/WhisperKit/commit/65cb888) |
+| [base.en](https://hf.co/argmaxinc/whisperkit-coreml/tree/main/openai_whisper-base.en)                 | [15.34](https://hf.co/datasets/argmaxinc/whisperkit-evals-staging/tree/main/WhisperKit/openai_whisper-base.en/earnings22-12hours)         |       7.7 |              145 | [Link](https://github.com/argmaxinc/WhisperKit/commit/65cb888) |
+| [base](https://hf.co/argmaxinc/whisperkit-coreml/tree/main/openai_whisper-base)                       | [16.62](https://hf.co/datasets/argmaxinc/whisperkit-evals-staging/tree/main/WhisperKit/openai_whisper-base/earnings22-12hours)            |       7.7 |              145 | [Link](https://github.com/argmaxinc/WhisperKit/commit/65cb888) |
+| [tiny.en](https://hf.co/argmaxinc/whisperkit-coreml/tree/main/openai_whisper-tiny.en)                 | [19.02](https://hf.co/datasets/argmaxinc/whisperkit-evals-staging/tree/main/WhisperKit/openai_whisper-tiny.en/earnings22-12hours)         |       0   |               66 | [Link](https://github.com/argmaxinc/WhisperKit/commit/65cb888) |
+| [tiny](https://hf.co/argmaxinc/whisperkit-coreml/tree/main/openai_whisper-tiny)                       | [21.21](https://hf.co/datasets/argmaxinc/whisperkit-evals-staging/tree/main/WhisperKit/openai_whisper-tiny/earnings22-12hours)            |       0   |               66 | [Link](https://github.com/argmaxinc/WhisperKit/commit/65cb888) |
+### Explanation
+We believe that rigorously measuring the quality of inference is necessary for developers and
+enterprises to make informed decisions when opting to use optimized or compressed variants of
+any machine learning model in production. To contextualize `WhisperKit`, we take the following Whisper
+implementations and benchmark them using a consistent evaluation harness:
+Server-side:
+- `WhisperOpenAIAPI`: [OpenAI's Whisper API](https://platform.openai.com/docs/guides/speech-to-text)
+($0.36 per hour of audio as of 02/29/24, 25MB file size limit per request)
+On-device:
+- `WhisperKit`: Argmax's implementation [[Eval Harness]](https://github.com/argmaxinc/whisperkittools/blob/main/whisperkit/pipelines.py#L100) [[Repo]](https://github.com/argmaxinc/WhisperKit)
+- `whisper.cpp`: A C++ implementation form ggerganov [[Eval Harness]](https://github.com/argmaxinc/whisperkittools/blob/main/whisperkit/pipelines.py#L212) [[Repo]](https://github.com/ggerganov/whisper.cpp)
+- `WhisperMLX`: A Python implementation from Apple MLX [[Eval Harness]](https://github.com/argmaxinc/whisperkittools/blob/main/whisperkit/pipelines.py#L338) [[Repo]](https://github.com/ml-explore/mlx-examples/blob/main/whisper/whisper/transcribe.py)
+(All on-device implementations are available for free under MIT license as of 03/19/2024)
+`WhisperOpenAIAPI` sets the reference and we assume that it is using the equivalent of [openai/whisper-large-v2](https://huggingface.co/openai/whisper-large-v2)
+in float16 precision along with additional undisclosed optimizations from OpenAI. In all measurements, we care primarily about per-example no-regressions (quantified as `qoi` below)
+which is a stricter metric compared to dataset average [Word Error RATE (WER)](https://en.wikipedia.org/wiki/Word_error_rate). A 100% `qoi` preserves perfect backwards-compatibility on the test distribution and avoids "perceived regressions", the phenomenon
+where per-example known behavior changes after a code/model update and causes divergence in downstream code or breaks the user experience itself (even if dataset averages might stay flat
+across updates). Pseudocode for `qoi`:
+```python
+qoi = []
+for example in dataset:
+    no_regression = wer(optimized_model(example)) <= wer(reference_model(example))
+    qoi.append(no_regression)
+qoi = (sum(qoi) / len(qoi)) * 100.
+```
+Note that the ordering of models with respect to `WER` does not necessarily match the ordering with respect to `QoI`. This is because the reference model gets assigned
+a QoI of 100% by definition. Any per-example regression by other implementations get penalized while per-example improvements are not rewarded. `QoI` (higher is better) matters
+where the production behavior is established by the reference results and the goal is to not regress when switching to an optimized or compressed model. On the other hand,
+`WER` (lower is better) matters when there is no established production behavior and one is picking the best quality versus model size trade off point.
+We anticipate developers that use Whisper (or similar models) in production to have their own Quality Assurance test sets and [whisperkittools](https://github.com/argmaxinc/whisperkittools) offers
+the tooling necessary to run the same measurements on such custom test sets, please see the [Model Evaluation on Custom Dataset]((https://github.com/argmaxinc/whisperkittools)) for details.
+### Why are there so many Whisper versions?
+WhisperKit is an SDK for building speech-to-text features in apps across a wide range of Apple devices. We are working towards abstracting away the model versioning from the developer so WhisperKit
+"just works" by deploying the highest-quality model version that a particular device can execute. In the interim, we leave the choice to the developer by providing quality and size trade-offs.
+### Datasets
+- [librispeech](https://huggingface.co/datasets/argmaxinc/librispeech): ~5 hours of short English audio clips, tests short-form transcription quality
+- [earnings22](https://huggingface.co/datasets/argmaxinc/earnings22): ~120 hours of English audio clips from earnings calls with various accents, tests long-form transcription quality
+### Reproducing Results
+Benchmark results on this page were automatically generated by [whisperkittools](https://github.com/argmaxinc/whisperkittools) using our cluster of Apple Silicon Macs as self-hosted runners on
+Github Actions. We periodically recompute these benchmarks as part of our CI pipeline. Due to [security concerns](https://docs.github.com/en/actions/security-guides/security-hardening-for-github-actions#hardening-for-self-hosted-runners),
+we are unable to open up the cluster to the public. However, any Apple Silicon Mac (even with 8GB RAM) can be used to
+run identical [evaluation jobs](#evaluation) locally. For reference, our M2 Ultra devices complete a `librispeech` + `openai/whisper-large-v3`
+evaluation in under 1 hour regardless of the Whisper implementation. Oldest Apple Silicon Macs should take less than 1 day to complete the same evaluation.
+### Glossary
+- `_turbo`: Indicates the presence of additional optimizations (not compression) to unlock streaming transcription
+as described in our [Blog Post](https://www.takeargmax.com/blog/whisperkit).
+- `_*MB`: Indicates the presence of model compression. Instead of cluttering the filename with details like
+`_AudioEncoder-5.8bits_TextDecoder-6.1bits_QLoRA-rank=16`, we choose to summarize the compression spec as the
+resulting total file size since this is what matters to developers in production.