Skip to content

vllm.v1.sample.sampler

A layer that samples the next tokens from the model's outputs.

Sampler

Bases: Module

A layer that samples the next tokens from the model's outputs with the following steps in order:

  1. If logprobs are requested: a) If logprobs_mode is raw_logprobs, compute logprobs as the final logprobs to return. b) If logprobs_mode is raw_logits, clone the logits as the final logprobs to return.
  2. Convert logits to float32.
  3. Apply allowed token ids whitelist.
  4. Apply bad words exclusion.
  5. Apply logit processors which are not argmax-invariant, i.e. that can impact greedy sampling. a) Min tokens processor b) Logit bias processor
  6. Apply penalties a) Repetition penalty b) Frequency penalty c) Presence penalty
  7. Sample the next tokens. sample method performs the following steps: a) If not all_random, perform greedy sampling. If all_greedy, return the greedily sampled tokens and final logprobs if requested. b) Apply temperature. c) Apply logit processors which are argmax-invariant, by default the min_p processor. d) Apply top_k and/or top_p. e) Sample the next tokens with the probability distribution. f) If all_random or temperature >= epsilon (1e-5), return the randomly sampled tokens and final logprobs if requested. Else, return the greedily sampled tokens and logprobs if requested.
  8. Gather the logprobs of the top max_num_logprobs and sampled token (if requested). Note that if the sampled token is within the top max_num_logprobs, the logprob will be eventually merged in LogprobsProcessor during output processing. Therefore, the final output may contain either max_num_logprobs + 1 or max_num_logprobs logprobs.
  9. Return the final SamplerOutput.
Source code in vllm/v1/sample/sampler.py
class Sampler(nn.Module):
    """
    A layer that samples the next tokens from the model's outputs
    with the following steps in order:

    1. If logprobs are requested:
        a) If `logprobs_mode` is `raw_logprobs`, compute logprobs
           as the final logprobs to return.
        b) If `logprobs_mode` is `raw_logits`, clone the logits
           as the final logprobs to return.
    2. Convert logits to float32.
    3. Apply allowed token ids whitelist.
    4. Apply bad words exclusion.
    5. Apply logit processors which are not argmax-invariant,
       i.e. that can impact greedy sampling.
        a) Min tokens processor
        b) Logit bias processor
    6. Apply penalties
        a) Repetition penalty
        b) Frequency penalty
        c) Presence penalty
    7. Sample the next tokens. `sample` method performs the following steps:
        a) If not `all_random`, perform greedy sampling. If `all_greedy`,
           return the greedily sampled tokens and final logprobs if requested.
        b) Apply temperature.
        c) Apply logit processors which are argmax-invariant, by default
           the min_p processor.
        d) Apply top_k and/or top_p.
        e) Sample the next tokens with the probability distribution.
        f) If `all_random` or temperature >= epsilon (1e-5), return the
           randomly sampled tokens and final logprobs if requested. Else,
           return the greedily sampled tokens and logprobs if requested.
    8. Gather the logprobs of the top `max_num_logprobs` and sampled token
       (if requested). Note that if the sampled token is within the top
       `max_num_logprobs`, the logprob will be eventually merged in
       `LogprobsProcessor` during output processing. Therefore, the
       final output may contain either `max_num_logprobs + 1` or
       `max_num_logprobs` logprobs.
    9. Return the final `SamplerOutput`.
    """

    def __init__(self, logprobs_mode: LogprobsMode = "raw_logprobs"):
        super().__init__()
        self.topk_topp_sampler = TopKTopPSampler(logprobs_mode)
        self.pin_memory = is_pin_memory_available()
        self.logprobs_mode = logprobs_mode

    def forward(
        self,
        logits: torch.Tensor,
        sampling_metadata: SamplingMetadata,
        predict_bonus_token: bool = False,
        logprobs_mode_override: LogprobsMode | None = None,
    ) -> SamplerOutput:
        logprobs_mode = logprobs_mode_override or self.logprobs_mode
        # NOTE(woosuk): Use the original logits (before any penalties or
        # temperature scaling) for the top-k logprobs.
        # This is different from the V0 sampler, which uses the logits that
        # is used for sampling (after penalties and temperature scaling).
        num_logprobs = sampling_metadata.max_num_logprobs
        if num_logprobs is not None:
            if logprobs_mode == "raw_logprobs":
                raw_logprobs = self.compute_logprobs(logits)
            elif logprobs_mode == "raw_logits":
                if logits.dtype == torch.float32:
                    raw_logprobs = logits.clone()
                else:
                    raw_logprobs = logits.to(torch.float32)

        # Use float32 for the logits.
        logits = logits.to(torch.float32)

        logits = self.apply_logits_processors(
            logits, sampling_metadata, predict_bonus_token
        )
        # Sample the next token.
        sampled, processed_logprobs = self.sample(logits, sampling_metadata)
        if processed_logprobs is not None:
            raw_logprobs = processed_logprobs
        # Convert sampled token ids to int64 (long) type to ensure compatibility
        # with subsequent operations that may use these values as indices.
        # This conversion is necessary because FlashInfer sampling operations
        # return int32 (while PyTorch argmax and topk return int64).
        sampled = sampled.long()

        # Handle logprob_token_ids if specified (more efficient than full vocab)
        # This is used by generative_scoring API to get logprobs for specific tokens
        logprob_token_ids_tensors = None
        if sampling_metadata.logprob_token_ids:
            logprob_token_ids_tensors = self.gather_specific_token_logprobs(
                logits, sampling_metadata.logprob_token_ids, sampled
            )

        if num_logprobs is None:
            logprobs_tensors = logprob_token_ids_tensors
        elif num_logprobs == -1:
            # Return the full unsorted and unranked logprobs.
            logprobs_tensors = LogprobsTensors(
                torch.empty(0), raw_logprobs, torch.empty(0)
            )
        else:
            # Gather the logprobs and ranks of the topk and sampled token.
            logprobs_tensors = self.gather_logprobs(
                raw_logprobs, num_logprobs, token_ids=sampled
            )

        # If we have both num_logprobs and logprob_token_ids, prefer
        # logprob_token_ids as it's more specific
        if logprob_token_ids_tensors is not None and num_logprobs is not None:
            logprobs_tensors = logprob_token_ids_tensors

        # Use int32 to reduce the tensor size.
        sampled = sampled.to(torch.int32)

        # These are GPU tensors.
        sampler_output = SamplerOutput(
            # The sampled tokens are expanded to 2D tensor with shape
            # [num_requests, 1], where each row represents one generated
            # token per request.
            sampled_token_ids=sampled.unsqueeze(-1),
            logprobs_tensors=logprobs_tensors,
        )
        return sampler_output

    def gather_specific_token_logprobs(
        self,
        logits: torch.Tensor,
        logprob_token_ids: dict[int, list[int]],
        sampled: torch.Tensor,
    ) -> LogprobsTensors | None:
        """Compute logprobs for specific token IDs using Triton kernel.

        This method handles heterogeneous token ID lists across requests by
        padding shorter lists to max length and using a fused Triton kernel
        for efficient log_softmax + gather computation.

        Benchmarks show the Triton kernel approach is ~1.4x faster than sparse
        gather for batch sizes > 1 due to the fused kernel reducing memory
        bandwidth requirements.

        Args:
            logits: [batch_size, vocab_size] tensor of logits
            logprob_token_ids: dict mapping req_index -> list of token IDs
            sampled: [batch_size] tensor of sampled token IDs

        Returns:
            LogprobsTensors with logprobs for the specified tokens, or None
            if no requests have logprob_token_ids.
        """
        if not logprob_token_ids:
            return None

        batch_size = logits.shape[0]
        device = logits.device

        # Find max number of tokens across all requests
        max_num_tokens = max(len(tids) for tids in logprob_token_ids.values())

        # Create padded token_ids tensor: [batch_size, max_num_tokens + 1]
        # +1 for sampled token in first position
        token_ids_tensor = torch.zeros(
            batch_size, max_num_tokens + 1, dtype=torch.int64, device=device
        )
        token_ids_tensor[:, 0] = sampled  # First column is sampled token

        # Create mask for valid positions (True = valid, False = padded)
        valid_mask = torch.zeros(
            batch_size, max_num_tokens + 1, dtype=torch.bool, device=device
        )
        valid_mask[:, 0] = True  # Sampled token is always valid

        # Fill in token IDs for each request
        for req_idx, token_ids in logprob_token_ids.items():
            num_tokens = len(token_ids)
            token_ids_tensor[req_idx, 1 : num_tokens + 1] = torch.tensor(
                token_ids, dtype=torch.int64, device=device
            )
            valid_mask[req_idx, 1 : num_tokens + 1] = True

        # Compute logprobs using the fused Triton kernel (log_softmax + gather)
        logprobs = compute_token_logprobs(logits, token_ids_tensor)

        # Mask invalid (padded) positions with -inf
        logprobs = logprobs.masked_fill(~valid_mask, float("-inf"))

        # Compute ranks for the sampled token
        sampled_logits = logits.gather(-1, sampled.unsqueeze(-1))
        token_ranks = (logits > sampled_logits).sum(dim=-1)

        return LogprobsTensors(
            logprob_token_ids=token_ids_tensor.to(torch.int32),
            logprobs=logprobs,
            selected_token_ranks=token_ranks,
        )

    @staticmethod
    def apply_temperature(
        logits: torch.Tensor,
        temp: torch.Tensor,
        all_random: bool,
    ) -> torch.Tensor:
        # Use in-place division to avoid creating a new tensor.
        # Avoid division by zero if there are greedy requests.
        if not all_random:
            temp = torch.where(temp < _SAMPLING_EPS, 1.0, temp)
        return logits.div_(temp.unsqueeze(dim=1))

    @staticmethod
    def greedy_sample(logits: torch.Tensor) -> torch.Tensor:
        return logits.argmax(dim=-1).view(-1)

    def sample(
        self,
        logits: torch.Tensor,
        sampling_metadata: SamplingMetadata,
        logprobs_mode_override: LogprobsMode | None = None,
    ) -> tuple[torch.Tensor, torch.Tensor | None]:
        """Sample logits based on sampling metadata.

        The various logits processing functions called in this method
        may update the logits tensor in-place.
        """

        logprobs_mode = logprobs_mode_override or self.logprobs_mode
        assert not (sampling_metadata.all_greedy and sampling_metadata.all_random)
        if sampling_metadata.all_random:
            greedy_sampled = None
        else:
            greedy_sampled = self.greedy_sample(logits)
            if sampling_metadata.all_greedy:
                processed_logprobs = None
                if sampling_metadata.max_num_logprobs is not None:
                    if logprobs_mode == "processed_logits":
                        processed_logprobs = logits
                    elif logprobs_mode == "processed_logprobs":
                        processed_logprobs = self.compute_logprobs(logits)
                return greedy_sampled, processed_logprobs

        assert sampling_metadata.temperature is not None

        # Apply temperature.
        logits = self.apply_temperature(
            logits, sampling_metadata.temperature, sampling_metadata.all_random
        )

        # Apply logits processors that only apply to random sampling
        # (argmax invariant)
        for processor in sampling_metadata.logitsprocs.argmax_invariant:
            logits = processor.apply(logits)

        # Apply top_k and/or top_p.
        random_sampled, processed_logprobs = self.topk_topp_sampler(
            logits,
            sampling_metadata.generators,
            sampling_metadata.top_k,
            sampling_metadata.top_p,
        )

        if greedy_sampled is None:
            return random_sampled, processed_logprobs

        sampled = torch.where(
            sampling_metadata.temperature < _SAMPLING_EPS,
            greedy_sampled,
            random_sampled,
            out=greedy_sampled,  # Reuse tensor
        )
        return sampled, processed_logprobs

    @staticmethod
    def compute_logprobs(logits: torch.Tensor) -> torch.Tensor:
        return logits.log_softmax(dim=-1, dtype=torch.float32)

    @staticmethod
    def gather_logprobs(
        logprobs: torch.Tensor,
        num_logprobs: int,
        token_ids: torch.Tensor,
    ) -> LogprobsTensors:
        """
        Gather logprobs for topk and sampled/prompt token.

        Args:
          logprobs: (num tokens) x (vocab) tensor
          num_logprobs: maximum number of logprobs to
                        retain per token
          token_ids: prompt tokens (if prompt logprobs)
                     or sampled tokens (if sampled
                     logprobs); 1D token ID tensor
                     with (num tokens) elements
                     Must be int64.

        Returns:
          Top-k int indices tensor, (num tokens) x (num_logprobs + 1)
          Top-k float logprobs tensor, (num tokens) x (num_logprobs + 1)
          Sampled token rank tensor, (num tokens)
        """
        assert token_ids.dtype == torch.int64
        # Find the topK values.
        topk_logprobs, topk_indices = torch.topk(logprobs, num_logprobs, dim=-1)

        # Get with the logprob of the prompt or sampled token.
        token_ids = token_ids.unsqueeze(-1)
        token_logprobs = logprobs.gather(-1, token_ids)

        # Compute the ranks of the actual token.
        token_ranks = batched_count_greater_than(logprobs, token_logprobs)

        # Concatenate together with the topk.
        indices = torch.cat((token_ids, topk_indices), dim=1)
        logprobs = torch.cat((token_logprobs, topk_logprobs), dim=1)

        # Use int32 to reduce the tensor size.
        indices = indices.to(torch.int32)

        return LogprobsTensors(indices, logprobs, token_ranks)

    @staticmethod
    def _combine_outputs_with_spec_tokens(
        output_token_ids: list[list[int]],
        spec_token_ids: list[list[int]] | None = None,
    ) -> list[list[int]]:
        if spec_token_ids is None:
            return output_token_ids

        return [
            [*out, *spec] if spec else out
            for out, spec in zip(output_token_ids, spec_token_ids)
        ]

    def apply_logits_processors(
        self,
        logits: torch.Tensor,
        sampling_metadata: SamplingMetadata,
        predict_bonus_token: bool,
    ) -> torch.Tensor:
        bad_words_token_ids = sampling_metadata.bad_words_token_ids
        any_penalties_or_bad_words = (
            bool(bad_words_token_ids) or not sampling_metadata.no_penalties
        )

        output_token_ids = sampling_metadata.output_token_ids
        if predict_bonus_token and any_penalties_or_bad_words:
            # Combine base outputs with spec tokens when speculative decoding
            # is enabled.
            output_token_ids = self._combine_outputs_with_spec_tokens(
                output_token_ids,
                sampling_metadata.spec_token_ids,
            )

        # Apply allowed token ids.
        if sampling_metadata.allowed_token_ids_mask is not None:
            logits.masked_fill_(sampling_metadata.allowed_token_ids_mask, float("-inf"))

        # Apply bad words exclusion.
        if bad_words_token_ids:
            apply_bad_words(logits, bad_words_token_ids, output_token_ids)

        # Apply logits processors which can impact greedy sampling.
        for processor in sampling_metadata.logitsprocs.non_argmax_invariant:
            logits = processor.apply(logits)

        # Apply penalties (e.g., freq_penalties).
        logits = self.apply_penalties(logits, sampling_metadata, output_token_ids)
        return logits

    @staticmethod
    def apply_penalties(
        logits: torch.Tensor,
        sampling_metadata: SamplingMetadata,
        output_token_ids: list[list[int]],
    ) -> torch.Tensor:
        if sampling_metadata.no_penalties:
            return logits

        assert sampling_metadata.prompt_token_ids is not None
        return apply_all_penalties(
            logits,
            sampling_metadata.prompt_token_ids,
            sampling_metadata.presence_penalties,
            sampling_metadata.frequency_penalties,
            sampling_metadata.repetition_penalties,
            output_token_ids,
        )

gather_logprobs staticmethod

gather_logprobs(
    logprobs: Tensor, num_logprobs: int, token_ids: Tensor
) -> LogprobsTensors

Gather logprobs for topk and sampled/prompt token.

Parameters:

Name Type Description Default
logprobs Tensor

(num tokens) x (vocab) tensor

required
num_logprobs int

maximum number of logprobs to retain per token

required
token_ids Tensor

prompt tokens (if prompt logprobs) or sampled tokens (if sampled logprobs); 1D token ID tensor with (num tokens) elements Must be int64.

required

Returns:

Type Description
LogprobsTensors

Top-k int indices tensor, (num tokens) x (num_logprobs + 1)

LogprobsTensors

Top-k float logprobs tensor, (num tokens) x (num_logprobs + 1)

LogprobsTensors

Sampled token rank tensor, (num tokens)

Source code in vllm/v1/sample/sampler.py
@staticmethod
def gather_logprobs(
    logprobs: torch.Tensor,
    num_logprobs: int,
    token_ids: torch.Tensor,
) -> LogprobsTensors:
    """
    Gather logprobs for topk and sampled/prompt token.

    Args:
      logprobs: (num tokens) x (vocab) tensor
      num_logprobs: maximum number of logprobs to
                    retain per token
      token_ids: prompt tokens (if prompt logprobs)
                 or sampled tokens (if sampled
                 logprobs); 1D token ID tensor
                 with (num tokens) elements
                 Must be int64.

    Returns:
      Top-k int indices tensor, (num tokens) x (num_logprobs + 1)
      Top-k float logprobs tensor, (num tokens) x (num_logprobs + 1)
      Sampled token rank tensor, (num tokens)
    """
    assert token_ids.dtype == torch.int64
    # Find the topK values.
    topk_logprobs, topk_indices = torch.topk(logprobs, num_logprobs, dim=-1)

    # Get with the logprob of the prompt or sampled token.
    token_ids = token_ids.unsqueeze(-1)
    token_logprobs = logprobs.gather(-1, token_ids)

    # Compute the ranks of the actual token.
    token_ranks = batched_count_greater_than(logprobs, token_logprobs)

    # Concatenate together with the topk.
    indices = torch.cat((token_ids, topk_indices), dim=1)
    logprobs = torch.cat((token_logprobs, topk_logprobs), dim=1)

    # Use int32 to reduce the tensor size.
    indices = indices.to(torch.int32)

    return LogprobsTensors(indices, logprobs, token_ranks)

gather_specific_token_logprobs

gather_specific_token_logprobs(
    logits: Tensor,
    logprob_token_ids: dict[int, list[int]],
    sampled: Tensor,
) -> LogprobsTensors | None

Compute logprobs for specific token IDs using Triton kernel.

This method handles heterogeneous token ID lists across requests by padding shorter lists to max length and using a fused Triton kernel for efficient log_softmax + gather computation.

Benchmarks show the Triton kernel approach is ~1.4x faster than sparse gather for batch sizes > 1 due to the fused kernel reducing memory bandwidth requirements.

Parameters:

Name Type Description Default
logits Tensor

[batch_size, vocab_size] tensor of logits

required
logprob_token_ids dict[int, list[int]]

dict mapping req_index -> list of token IDs

required
sampled Tensor

[batch_size] tensor of sampled token IDs

required

Returns:

Type Description
LogprobsTensors | None

LogprobsTensors with logprobs for the specified tokens, or None

LogprobsTensors | None

if no requests have logprob_token_ids.

Source code in vllm/v1/sample/sampler.py
def gather_specific_token_logprobs(
    self,
    logits: torch.Tensor,
    logprob_token_ids: dict[int, list[int]],
    sampled: torch.Tensor,
) -> LogprobsTensors | None:
    """Compute logprobs for specific token IDs using Triton kernel.

    This method handles heterogeneous token ID lists across requests by
    padding shorter lists to max length and using a fused Triton kernel
    for efficient log_softmax + gather computation.

    Benchmarks show the Triton kernel approach is ~1.4x faster than sparse
    gather for batch sizes > 1 due to the fused kernel reducing memory
    bandwidth requirements.

    Args:
        logits: [batch_size, vocab_size] tensor of logits
        logprob_token_ids: dict mapping req_index -> list of token IDs
        sampled: [batch_size] tensor of sampled token IDs

    Returns:
        LogprobsTensors with logprobs for the specified tokens, or None
        if no requests have logprob_token_ids.
    """
    if not logprob_token_ids:
        return None

    batch_size = logits.shape[0]
    device = logits.device

    # Find max number of tokens across all requests
    max_num_tokens = max(len(tids) for tids in logprob_token_ids.values())

    # Create padded token_ids tensor: [batch_size, max_num_tokens + 1]
    # +1 for sampled token in first position
    token_ids_tensor = torch.zeros(
        batch_size, max_num_tokens + 1, dtype=torch.int64, device=device
    )
    token_ids_tensor[:, 0] = sampled  # First column is sampled token

    # Create mask for valid positions (True = valid, False = padded)
    valid_mask = torch.zeros(
        batch_size, max_num_tokens + 1, dtype=torch.bool, device=device
    )
    valid_mask[:, 0] = True  # Sampled token is always valid

    # Fill in token IDs for each request
    for req_idx, token_ids in logprob_token_ids.items():
        num_tokens = len(token_ids)
        token_ids_tensor[req_idx, 1 : num_tokens + 1] = torch.tensor(
            token_ids, dtype=torch.int64, device=device
        )
        valid_mask[req_idx, 1 : num_tokens + 1] = True

    # Compute logprobs using the fused Triton kernel (log_softmax + gather)
    logprobs = compute_token_logprobs(logits, token_ids_tensor)

    # Mask invalid (padded) positions with -inf
    logprobs = logprobs.masked_fill(~valid_mask, float("-inf"))

    # Compute ranks for the sampled token
    sampled_logits = logits.gather(-1, sampled.unsqueeze(-1))
    token_ranks = (logits > sampled_logits).sum(dim=-1)

    return LogprobsTensors(
        logprob_token_ids=token_ids_tensor.to(torch.int32),
        logprobs=logprobs,
        selected_token_ranks=token_ranks,
    )

sample

sample(
    logits: Tensor,
    sampling_metadata: SamplingMetadata,
    logprobs_mode_override: LogprobsMode | None = None,
) -> tuple[Tensor, Tensor | None]

Sample logits based on sampling metadata.

The various logits processing functions called in this method may update the logits tensor in-place.

Source code in vllm/v1/sample/sampler.py
def sample(
    self,
    logits: torch.Tensor,
    sampling_metadata: SamplingMetadata,
    logprobs_mode_override: LogprobsMode | None = None,
) -> tuple[torch.Tensor, torch.Tensor | None]:
    """Sample logits based on sampling metadata.

    The various logits processing functions called in this method
    may update the logits tensor in-place.
    """

    logprobs_mode = logprobs_mode_override or self.logprobs_mode
    assert not (sampling_metadata.all_greedy and sampling_metadata.all_random)
    if sampling_metadata.all_random:
        greedy_sampled = None
    else:
        greedy_sampled = self.greedy_sample(logits)
        if sampling_metadata.all_greedy:
            processed_logprobs = None
            if sampling_metadata.max_num_logprobs is not None:
                if logprobs_mode == "processed_logits":
                    processed_logprobs = logits
                elif logprobs_mode == "processed_logprobs":
                    processed_logprobs = self.compute_logprobs(logits)
            return greedy_sampled, processed_logprobs

    assert sampling_metadata.temperature is not None

    # Apply temperature.
    logits = self.apply_temperature(
        logits, sampling_metadata.temperature, sampling_metadata.all_random
    )

    # Apply logits processors that only apply to random sampling
    # (argmax invariant)
    for processor in sampling_metadata.logitsprocs.argmax_invariant:
        logits = processor.apply(logits)

    # Apply top_k and/or top_p.
    random_sampled, processed_logprobs = self.topk_topp_sampler(
        logits,
        sampling_metadata.generators,
        sampling_metadata.top_k,
        sampling_metadata.top_p,
    )

    if greedy_sampled is None:
        return random_sampled, processed_logprobs

    sampled = torch.where(
        sampling_metadata.temperature < _SAMPLING_EPS,
        greedy_sampled,
        random_sampled,
        out=greedy_sampled,  # Reuse tensor
    )
    return sampled, processed_logprobs