sampler.py

# SPDX-License-Identifier: Apache-2.0
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
from collections.abc import Callable

import torch

from vllm.config.model import LogprobsMode
from vllm.triton_utils import tl, triton
from vllm.v1.outputs import LogprobsTensors, SamplerOutput
from vllm.v1.sample.ops.topk_topp_sampler import apply_top_k_top_p
from vllm.v1.worker.gpu.states import SamplingMetadata


class Sampler:
    def __init__(
        self,
        logprobs_mode: LogprobsMode = "raw_logprobs",
    ):
        if logprobs_mode not in ["processed_logprobs", "raw_logprobs"]:
            raise NotImplementedError(f"Unsupported logprobs_mode: {logprobs_mode}")
        self.logprobs_mode = logprobs_mode

    def __call__(
        self,
        logits: torch.Tensor,
        sampling_metadata: SamplingMetadata,
    ) -> SamplerOutput:
        if sampling_metadata.max_num_logprobs is not None:
            if self.logprobs_mode == "processed_logprobs":
                sampled, logits = self.sample(
                    logits, sampling_metadata, return_logits=True
                )
            else:
                assert self.logprobs_mode == "raw_logprobs"
                sampled, _ = self.sample(logits, sampling_metadata, return_logits=False)

            logprobs_tensors = compute_topk_logprobs(
                logits,
                sampling_metadata.max_num_logprobs,
                sampled,
            )
        else:
            sampled, _ = self.sample(logits, sampling_metadata, return_logits=False)
            logprobs_tensors = None

        # 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.view(-1, 1),
            logprobs_tensors=logprobs_tensors,
        )
        return sampler_output

    def sample(
        self,
        logits: torch.Tensor,
        sampling_metadata: SamplingMetadata,
        return_logits: bool = False,
    ) -> tuple[torch.Tensor, torch.Tensor | None]:
        is_greedy = sampling_metadata.temperature == 0
        temp = torch.where(is_greedy, 1.0, sampling_metadata.temperature)
        logits = logits / temp.view(-1, 1)
        logits = apply_top_k_top_p(
            logits, sampling_metadata.top_k, sampling_metadata.top_p
        )

        sampled = gumbel_sample(
            logits,
            is_greedy,
            sampling_metadata.seeds,
            sampling_metadata.pos,
        )
        return sampled, logits if return_logits else None


@triton.jit
def _gumbel_sample_kernel(
    local_argmax_ptr,
    local_argmax_stride,
    local_max_ptr,
    local_max_stride,
    logits_ptr,
    logits_stride,
    seeds_ptr,
    pos_ptr,
    is_greedy_ptr,
    vocab_size,
    BLOCK_SIZE: tl.constexpr,
):
    req_idx = tl.program_id(0)
    block_idx = tl.program_id(1)
    block = block_idx * BLOCK_SIZE + tl.arange(0, BLOCK_SIZE)
    mask = block < vocab_size
    logits = tl.load(
        logits_ptr + req_idx * logits_stride + block,
        mask=mask,
        other=float("-inf"),
    )

    is_greedy = tl.load(is_greedy_ptr + req_idx)
    if not is_greedy:
        # Calculate the seed for gumbel noise.
        seed = tl.load(seeds_ptr + req_idx)
        pos = tl.load(pos_ptr + req_idx)
        gumbel_seed = tl.randint(seed, pos)

        # Generate gumbel noise.
        r = tl.rand(gumbel_seed, block).to(tl.float64)
        gumbel_noise = -tl.log(-tl.log(r + 1e-20) + 1e-20)
        gumbel_noise = gumbel_noise.to(tl.float32)

        # Apply gumbel noise.
        logits = tl.where(mask, logits + gumbel_noise, float("-inf"))

    idx = tl.argmax(logits, axis=0)
    token_id = block_idx * BLOCK_SIZE + idx
    value = tl.max(logits, axis=0)
    tl.store(local_argmax_ptr + req_idx * local_argmax_stride + block_idx, token_id)
    tl.store(local_max_ptr + req_idx * local_max_stride + block_idx, value)


def gumbel_sample(
    logits: torch.Tensor,  # [num_reqs, vocab_size]
    is_greedy: torch.Tensor,  # [num_reqs]
    seed: torch.Tensor,  # [num_reqs]
    pos: torch.Tensor,  # [num_reqs]
) -> torch.Tensor:
    num_reqs, vocab_size = logits.shape
    BLOCK_SIZE = 1024
    num_blocks = triton.cdiv(vocab_size, BLOCK_SIZE)
    local_argmax = torch.empty(
        num_reqs,
        num_blocks,
        dtype=torch.int64,
        device=logits.device,
    )
    local_max = torch.empty(
        num_reqs,
        num_blocks,
        dtype=torch.float32,
        device=logits.device,
    )
    _gumbel_sample_kernel[(num_reqs, num_blocks)](
        local_argmax,
        local_argmax.stride(0),
        local_max,
        local_max.stride(0),
        logits,
        logits.stride(0),
        seed,
        pos,
        is_greedy,
        vocab_size,
        BLOCK_SIZE=BLOCK_SIZE,
    )
    # NOTE(woosuk): Use int64 for later indexing.
    max_block_idx = local_max.argmax(dim=-1, keepdim=True)
    sampled = local_argmax.gather(dim=-1, index=max_block_idx).view(-1)
    return sampled


@triton.jit
def _topk_log_softmax_kernel(
    output_ptr,
    logits_ptr,
    logits_stride,
    topk_ids_ptr,
    topk,
    vocab_size,
    BLOCK_SIZE: tl.constexpr,
    PADDED_TOPK: tl.constexpr,
):
    req_idx = tl.program_id(0)
    row_ptr = logits_ptr + req_idx * logits_stride

    max_val = float("-inf")
    for i in range(0, vocab_size, BLOCK_SIZE):
        block = i + tl.arange(0, BLOCK_SIZE)
        logits = tl.load(row_ptr + block, mask=block < vocab_size, other=float("-inf"))
        max_val = tl.max(tl.maximum(logits, max_val))
    max_val = max_val.to(tl.float32)  # type: ignore

    se = 0.0
    for i in range(0, vocab_size, BLOCK_SIZE):
        block = i + tl.arange(0, BLOCK_SIZE)
        logits = tl.load(row_ptr + block, mask=block < vocab_size, other=0.0)
        # NOTE(woosuk): Make sure that logits and all following operations use FP32.
        logits = logits.to(tl.float32)
        e = tl.exp(logits - max_val)
        e = tl.where(block < vocab_size, e, 0.0)
        se += tl.sum(e)
    lse = tl.log(se)

    k_offset = tl.arange(0, PADDED_TOPK)
    k_mask = k_offset < topk
    topk_ids = tl.load(topk_ids_ptr + req_idx * topk + k_offset, mask=k_mask, other=0)

    logits = tl.load(row_ptr + topk_ids, mask=k_mask)
    logits = logits.to(tl.float32)
    o = logits - max_val - lse
    tl.store(output_ptr + req_idx * topk + k_offset, o, mask=k_mask)


@triton.jit
def _ranks_kernel(
    output_ptr,
    logits_ptr,
    logits_stride,
    token_ids_ptr,
    vocab_size,
    BLOCK_SIZE: tl.constexpr,
):
    req_idx = tl.program_id(0)
    row_ptr = logits_ptr + req_idx * logits_stride

    token_id = tl.load(token_ids_ptr + req_idx)
    x = tl.load(row_ptr + token_id)

    n = 0
    for i in range(0, vocab_size, BLOCK_SIZE):
        block = i + tl.arange(0, BLOCK_SIZE)
        logits = tl.load(row_ptr + block, mask=block < vocab_size, other=float("-inf"))
        n += tl.sum((logits > x).to(tl.int32))
    tl.store(output_ptr + req_idx, n)


def compute_token_logprobs(
    logits: torch.Tensor,
    token_ids: torch.Tensor,
) -> torch.Tensor:
    batch_size = logits.shape[0]
    vocab_size = logits.shape[1]
    token_ids = token_ids.to(torch.int64)
    num_logprobs = token_ids.shape[1]
    logprobs = torch.empty(
        batch_size,
        num_logprobs,
        dtype=torch.float32,
        device=logits.device,
    )
    _topk_log_softmax_kernel[(batch_size,)](
        logprobs,
        logits,
        logits.stride(0),
        token_ids,
        num_logprobs,
        vocab_size,
        BLOCK_SIZE=1024,  # type: ignore
        PADDED_TOPK=triton.next_power_of_2(num_logprobs),
    )
    return logprobs


def compute_topk_logprobs(
    logits: torch.Tensor,
    num_logprobs: int,
    sampled_token_ids: torch.Tensor,
) -> LogprobsTensors:
    assert num_logprobs >= 0
    batch_size, vocab_size = logits.shape
    if num_logprobs == 0:
        logprob_token_ids = sampled_token_ids.unsqueeze(-1)
    else:
        topk_indices = torch.topk(logits, num_logprobs, dim=-1).indices
        logprob_token_ids = torch.cat(
            (sampled_token_ids.unsqueeze(-1), topk_indices), dim=1
        )

    # NOTE(woosuk): Here, to save GPU memory, we do not materialize the full
    # logprobs tensor. Instead, we only compute and return the logprobs of
    # the topk + 1 tokens.
    logprobs = compute_token_logprobs(logits, logprob_token_ids)
    token_ranks = torch.empty(
        batch_size,
        dtype=torch.int64,
        device=logits.device,
    )
    _ranks_kernel[(batch_size,)](
        token_ranks,
        logits,
        logits.stride(0),
        sampled_token_ids,
        vocab_size,
        BLOCK_SIZE=8192,  # type: ignore
    )
    return LogprobsTensors(
        logprob_token_ids=logprob_token_ids,
        logprobs=logprobs,
        selected_token_ranks=token_ranks,
    )


def compute_prompt_logprobs(
    prompt_token_ids: torch.Tensor,
    prompt_hidden_states: torch.Tensor,
    logits_fn: Callable[[torch.Tensor], torch.Tensor],
) -> tuple[torch.Tensor, torch.Tensor]:
    # Since materializing the full prompt logits can take too much memory,
    # we compute it in chunks.
    CHUNK_SIZE = 1024
    logprobs = []
    ranks = []
    prompt_token_ids = prompt_token_ids.to(torch.int64)
    for start_idx in range(0, prompt_token_ids.shape[0], CHUNK_SIZE):
        end_idx = start_idx + CHUNK_SIZE
        # NOTE(woosuk): logits_fn can be slow because it involves all-gather.
        prompt_logits = logits_fn(prompt_hidden_states[start_idx:end_idx])
        prompt_logprobs = compute_topk_logprobs(
            prompt_logits,
            0,  # num_logprobs
            prompt_token_ids[start_idx:end_idx],
        )
        logprobs.append(prompt_logprobs.logprobs)
        ranks.append(prompt_logprobs.selected_token_ranks)

    logprobs = torch.cat(logprobs, dim=0) if len(logprobs) > 1 else logprobs[0]
    ranks = torch.cat(ranks, dim=0) if len(ranks) > 1 else ranks[0]
    return logprobs, ranks