vllm.benchmarks.datasets ¶

class AIMODataset(HuggingFaceDataset):
    """
    Dataset class for processing a AIMO dataset with reasoning questions.
    """
    SUPPORTED_DATASET_PATHS = {
        "AI-MO/aimo-validation-aime", "AI-MO/NuminaMath-1.5",
        "AI-MO/NuminaMath-CoT"
    }

    def sample(self,
               tokenizer: PreTrainedTokenizerBase,
               num_requests: int,
               output_len: Optional[int] = None,
               request_id_prefix: str = "",
               no_oversample: bool = False,
               **kwargs) -> list:
        sampled_requests = []
        ind = 0
        dynamic_output = output_len is None

        for item in self.data:
            if len(sampled_requests) >= num_requests:
                break
            prompt, completion = item['problem'], item["solution"]

            prompt_ids = tokenizer(prompt).input_ids
            completion_ids = tokenizer(completion).input_ids
            prompt_len = len(prompt_ids)
            completion_len = len(completion_ids)
            output_len = completion_len if dynamic_output else output_len
            assert isinstance(output_len, int) and output_len > 0
            if dynamic_output and not is_valid_sequence(prompt_len,
                                                        completion_len,
                                                        max_prompt_len=2048,
                                                        max_total_len=32000):
                continue
            sampled_requests.append(
                SampleRequest(
                    prompt=prompt,
                    prompt_len=prompt_len,
                    expected_output_len=output_len,
                    multi_modal_data=None,
                    request_id=request_id_prefix + str(ind),
                ))
            ind += 1
        self.maybe_oversample_requests(sampled_requests, num_requests,
                                       request_id_prefix, no_oversample)
        return sampled_requests

class ASRDataset(HuggingFaceDataset):
    """
    Dataset class for processing a ASR dataset for transcription.
    Tested on the following set:

    +----------------+----------------------------------------+--------------------------+-----------------------------+
    | Dataset        | Domain                                 | Speaking Style           | hf-subset                   |
    +----------------+----------------------------------------+--------------------------+-----------------------------+
    | TED-LIUM       | TED talks                              | Oratory                  | release1, release2, release3|
    |                |                                        |                          | release3-speaker-adaptation |
    | VoxPopuli      | European Parliament                    | Oratory                  | en, de, it, fr,  ...        |
    | LibriSpeech    | Audiobook                              | Narrated                 | "LIUM/tedlium"              |
    | GigaSpeech     | Audiobook, podcast, YouTube            | Narrated, spontaneous    | xs, s, m, l, xl, dev, test  |
    | SPGISpeech     | Financial meetings                     | Oratory, spontaneous     | S, M, L, dev, test          |
    | AMI            | Meetings                               | Spontaneous              | ihm, sdm                    |
    +----------------+----------------------------------------+--------------------------+-----------------------------+

    """  # noqa: E501

    SUPPORTED_DATASET_PATHS = {
        "openslr/librispeech_asr",
        "facebook/voxpopuli",
        "LIUM/tedlium",
        "edinburghcstr/ami",
        "speechcolab/gigaspeech",
        "kensho/spgispeech",
    }

    DEFAULT_OUTPUT_LEN = 128
    IS_MULTIMODAL = True

    # TODO Whisper-specific. Abstract interface when more models are supported.
    TRANSCRIPTION_PREAMBLE = (
        "<|startoftranscript|><|en|><|transcribe|><|notimestamps|>")
    skip_long_audios: bool = True

    def sample(
        self,
        tokenizer: PreTrainedTokenizerBase,
        num_requests: int,
        output_len: Optional[int] = None,
        request_id_prefix: str = "",
        no_oversample: bool = False,
        **kwargs,
    ) -> list:
        output_len = (output_len
                      if output_len is not None else self.DEFAULT_OUTPUT_LEN)
        prompt = ASRDataset.TRANSCRIPTION_PREAMBLE
        prompt_len = len(tokenizer(prompt).input_ids)
        sampled_requests = []
        ind = 0
        skipped = 0
        for item in self.data:
            if len(sampled_requests) >= num_requests:
                break
            audio = item["audio"]
            y, sr = audio["array"], audio["sampling_rate"]
            duration_s = librosa.get_duration(y=y, sr=sr)
            # Whisper max supported duration
            if self.skip_long_audios and duration_s > 30:
                skipped += 1
                continue

            mm_content = {"audio": (y, sr)}
            sampled_requests.append(
                SampleRequest(
                    prompt=prompt,
                    prompt_len=prompt_len,
                    expected_output_len=output_len,
                    multi_modal_data=mm_content,
                    request_id=request_id_prefix + str(ind),
                ))
            ind += 1
        if skipped:
            logger.warning(
                "%d samples discarded from dataset due to"
                " their length being greater than"
                " what Whisper supports.",
                skipped,
            )
        self.maybe_oversample_requests(sampled_requests, num_requests,
                                       request_id_prefix, no_oversample)
        return sampled_requests

class BenchmarkDataset(ABC):
    DEFAULT_SEED = 0
    IS_MULTIMODAL = False

    def __init__(
        self,
        dataset_path: Optional[str] = None,
        random_seed: int = DEFAULT_SEED,
    ) -> None:
        """
        Initialize the BenchmarkDataset with an optional dataset path and random
        seed.

        Args:
            dataset_path (Optional[str]): Path to the dataset. If None, it
                indicates that a default or random dataset might be used.
            random_seed (int): Seed value for reproducible shuffling or
                sampling. Defaults to DEFAULT_SEED.
        """
        self.dataset_path = dataset_path
        # Set the random seed, ensuring that a None value is replaced with the
        # default seed.
        self.random_seed = (random_seed
                            if random_seed is not None else self.DEFAULT_SEED)
        self.data = None

    def apply_multimodal_chat_transformation(
            self,
            prompt: str,
            mm_content: Optional[
                        Union[MultiModalDataDict, dict, list[dict]]
                             ] = None) -> list[dict]:
        """
        Transform a prompt and optional multimodal content into a chat format.
        This method is used for chat models that expect a specific conversation
        format.
        """
        content = [{"text": prompt, "type": "text"}]
        if mm_content is not None:
            if isinstance(mm_content, list):
                content.extend(cast(list[dict[str, Any]], mm_content))
            elif isinstance(mm_content, dict):
                content.append(mm_content)
            else:
                raise TypeError(
                    "Could not process multimodal content of type: " +
                    f"{type(mm_content)}"
                )
        return [{"role": "user", "content": content}]

    def load_data(self) -> None:
        """
        Load data from the dataset path into self.data.

        This method must be overridden by subclasses since the method to load
        data will vary depending on the dataset format and source.

        Raises:
            NotImplementedError: If a subclass does not implement this method.
        """
        # TODO (jenniferzhao): add support for downloading data
        raise NotImplementedError(
            "load_data must be implemented in subclasses.")

    def get_random_lora_request(
        self,
        max_loras: Optional[int] = None,
        lora_path: Optional[str] = None,
    ) -> Optional[LoRARequest]:
        """
        Optionally select a random LoRA request.

        This method is used when LoRA parameters are provided.  It randomly
        selects a LoRA based on max_loras.

        Args:
            max_loras (Optional[int]): The maximum number of LoRAs available.
                If `None`, LoRA is not used.
            lora_path (Optional[str]): Path to the LoRA parameters on disk.
                If `None`, LoRA is not used.

        Returns:
            A new [`LoRARequest`][vllm.lora.request.LoRARequest]
            (or `None` if not applicable).
        """
        if max_loras is None or lora_path is None:
            return None

        # Generate a random LoRA ID in the range [1, max_loras].
        lora_id = random.randint(1, max_loras)
        lora_request = LoRARequest(
            lora_name=str(lora_id),
            lora_int_id=lora_id,
            lora_path=lora_path_on_disk(lora_path),
        )
        return lora_request

    @abstractmethod
    def sample(self, tokenizer: PreTrainedTokenizerBase,
               num_requests: int,
               request_id_prefix: str = "",
               no_oversample: bool = False) -> list[SampleRequest]:
        """
        Abstract method to generate sample requests from the dataset.

        Subclasses must override this method to implement dataset-specific logic
        for generating a list of SampleRequest objects.

        Args:
            tokenizer (PreTrainedTokenizerBase): The tokenizer to be used
                for processing the dataset's text.
            num_requests (int): The number of sample requests to generate.
            request_id_prefix (str): The prefix of request_id.

        Returns:
            list[SampleRequest]: A list of sample requests generated from the
            dataset.
        """
        raise NotImplementedError("sample must be implemented in subclasses.")

    def maybe_oversample_requests(
        self,
        requests: list[SampleRequest],
        num_requests: int,
        request_id_prefix: str = "",
        no_oversample: bool = False,
    ) -> None:
        """
        Oversamples the list of requests if its size is less than the desired
        number.

        Args:
            requests (List[SampleRequest]): The current list of sampled
                requests.
            num_requests (int): The target number of requests.
            request_id_prefix (str): The prefix applied to generated request
                identifiers.

        """
        if no_oversample:
            logger.info("Skipping oversampling. " \
            "Total samples: %d.", len(requests))
            return

        if len(requests) < num_requests:
            random.seed(self.random_seed)
            additional = deepcopy(
                random.choices(requests, k=num_requests - len(requests))
            )
            for i in range(len(additional)):
                req = additional[i]
                req.request_id = request_id_prefix + str(len(requests) + i)
            requests.extend(additional)
            logger.info("Oversampled requests to reach %d total samples.",
                        num_requests)

class BlazeditDataset(HuggingFaceDataset):
    """
    Blazedit Dataset.
    https://github.com/ise-uiuc/blazedit

    5k char version: vdaita/edit_5k_char
    10k char version: vdaita/edit_10k_char
    """  # noqa: E501

    # 5k char version will have output as ~5k chars
    # 10k char version will have output as ~10k chars
    # Assuming 3 char per token, 10k chars will be 3333 tokens
    # We set default to 4000 to be safe
    DEFAULT_OUTPUT_LEN = 4000
    SUPPORTED_DATASET_PATHS = {
        "vdaita/edit_5k_char",
        "vdaita/edit_10k_char",
    }

    def sample(
        self,
        tokenizer: PreTrainedTokenizerBase,
        num_requests: int,
        output_len: Optional[int] = None,
        request_id_prefix: str = "",
        no_oversample: bool = False,
        min_distance: float = 0.0,
        max_distance: float = 1.0,
        **kwargs,
    ) -> list:
        output_len = (output_len
                      if output_len is not None else self.DEFAULT_OUTPUT_LEN)
        sampled_requests = []

        for i, item in enumerate(self.data):
            if len(sampled_requests) >= num_requests:
                break
            code = item["code"]
            change_request = item["change_request"]
            norm_distance = item["norm_distance"]

            # compare the levenshtein distance normalized by code length
            if norm_distance < min_distance or norm_distance > max_distance:
                continue

            # template copied from
            # https://github.com/ise-uiuc/blazedit/blob/7765137e656fd62de877422d2e4cf8de51228054/dataset/create_refined_dataset.py#L94-L105 # noqa: E501
            instruction = f"""Given a code file, please apply the change requests and generate the new file.

Original file:
```python
{code}
```

Change request:
{change_request}

Please generate the new code file in the "New file" section below.""" # noqa: E501

            # apply template
            prompt = tokenizer.apply_chat_template(
                [{
                    "role": "user",
                    "content": instruction
                }],
                add_generation_prompt=True,
                tokenize=False,
            )

            prompt_len = len(tokenizer(prompt).input_ids)

            sampled_requests.append(
                SampleRequest(
                    prompt=prompt,
                    prompt_len=prompt_len,
                    expected_output_len=output_len,
                    request_id=request_id_prefix + str(i),
                ))
        self.maybe_oversample_requests(sampled_requests, num_requests,
                                       request_id_prefix, no_oversample)

        return sampled_requests

class BurstGPTDataset(BenchmarkDataset):
    """
    Implements the BurstGPT dataset.  Loads data from a CSV file and generates
    sample requests based on synthetic prompt generation. Only rows with Model
    "GPT-4" and positive response tokens are used.
    """

    def __init__(self, **kwargs) -> None:
        super().__init__(**kwargs)
        self.load_data()

    def load_data(self, ):
        if self.dataset_path is None:
            raise ValueError("dataset_path must be provided for loading data.")

        df = pd.read_csv(self.dataset_path)
        # Filter to keep only GPT-4 rows.
        gpt4_df = df[df["Model"] == "GPT-4"]
        # Remove failed requests (where Response tokens is 0 or less).
        gpt4_df = gpt4_df[gpt4_df["Response tokens"] > 0]
        # Sample the desired number of rows.
        self.data = gpt4_df

    def _sample_loaded_data(self, num_requests: int) -> list:
        if num_requests <= len(self.data):
            data = self.data.sample(n=num_requests,
                                    random_state=self.random_seed)
        else:
            data = self.data.sample(
                n=num_requests,
                random_state=self.random_seed,
                replace=True,
            )
        # Convert the dataframe to a list of lists.
        return data.values.tolist()

    def sample(
        self,
        tokenizer: PreTrainedTokenizerBase,
        num_requests: int,
        max_loras: Optional[int] = None,
        lora_path: Optional[str] = None,
        request_id_prefix: str = "",
        no_oversample: bool = False,
        **kwargs,
    ) -> list[SampleRequest]:
        samples = []
        data = self._sample_loaded_data(num_requests=num_requests)
        for i in range(num_requests):
            input_len = int(data[i][2])
            output_len = int(data[i][3])
            lora_req = self.get_random_lora_request(
                max_loras=max_loras, lora_path=lora_path)
            vocab_size = tokenizer.vocab_size
            # Generate a synthetic prompt: a list of token IDs computed as (i +
            # j) modulo vocab_size.
            token_ids = [(i + j) % vocab_size for j in range(input_len)]
            prompt = tokenizer.decode(token_ids)
            samples.append(
                SampleRequest(
                    prompt=prompt,
                    prompt_len=input_len,
                    expected_output_len=output_len,
                    lora_request=lora_req,
                    request_id=request_id_prefix + str(i),
                ))
        return samples

class ConversationDataset(HuggingFaceDataset):
    """Dataset for conversation data with multimodal support."""
    SUPPORTED_DATASET_PATHS = {
        'lmms-lab/LLaVA-OneVision-Data', 'Aeala/ShareGPT_Vicuna_unfiltered'
    }
    IS_MULTIMODAL = True

    def sample(self,
               tokenizer: PreTrainedTokenizerBase,
               num_requests: int,
               output_len: Optional[int] = None,
               enable_multimodal_chat: bool = False,
               request_id_prefix: str = "",
               no_oversample: bool = False,
               **kwargs) -> list:
        # Filter examples with at least 2 conversations
        filtered_data = self.data.filter(
            lambda x: len(x["conversations"]) >= 2)
        sampled_requests = []
        ind = 0
        dynamic_output = output_len is None

        for item in filtered_data:
            if len(sampled_requests) >= num_requests:
                break
            conv = item["conversations"]
            prompt, completion = conv[0]["value"], conv[1]["value"]

            prompt_ids = tokenizer(prompt).input_ids
            completion_ids = tokenizer(completion).input_ids
            prompt_len = len(prompt_ids)
            completion_len = len(completion_ids)
            output_len = completion_len if dynamic_output else output_len
            assert isinstance(output_len, int) and output_len > 0
            if dynamic_output and not is_valid_sequence(
                    prompt_len, completion_len):
                continue
            mm_content = process_image(
                item["image"]) if "image" in item else None
            if enable_multimodal_chat:
                # Note: when chat is enabled the request prompt_len is no longer
                # accurate and we will be using request output to count the
                # actual prompt len and output len
                prompt = self.apply_multimodal_chat_transformation(
                    prompt, mm_content)
            sampled_requests.append(
                SampleRequest(
                    prompt=prompt,
                    prompt_len=prompt_len,
                    expected_output_len=output_len,
                    multi_modal_data=mm_content,
                    request_id=request_id_prefix + str(ind),
                ))
            ind += 1
        self.maybe_oversample_requests(sampled_requests, num_requests,
                                       request_id_prefix, no_oversample)
        return sampled_requests