[V1][Spec Decode] Implement Eagle Proposer [1/N] (#15729)

Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>

vllm/config.py

@@ -2152,9 +2152,10 @@ class SpeculativeConfig:
 
                 # Replace hf_config for EAGLE draft_model
                 if self.method == "eagle":
-                    if self.enable_chunked_prefill:
+                    if self.enable_chunked_prefill and not envs.VLLM_USE_V1:
                         raise ValueError(
-                            "Chunked prefill and EAGLE are not compatible.")
+                            "Chunked prefill and EAGLE are not compatible "
+                            "when using V0.")
 
                     from vllm.transformers_utils.configs.eagle import (
                         EAGLEConfig)

vllm/engine/arg_utils.py

@@ -1468,15 +1468,21 @@ class EngineArgs:
 
         # Only Ngram speculative decoding so far.
         is_ngram_enabled = False
+        is_eagle_enabled = False
         if self.speculative_config is not None:
             # This is supported but experimental (handled below).
-            if (("method" in self.speculative_config
-                 and self.speculative_config["method"] in ("ngram", "[ngram]"))
-                    or
-                ("model" in self.speculative_config and
-                 self.speculative_config["model"] in ("ngram", "[ngram]"))):
+            speculative_method = self.speculative_config.get("method")
+            if speculative_method:
+                if speculative_method in ("ngram", "[ngram]"):
                     is_ngram_enabled = True
+                elif speculative_method == "eagle":
+                    is_eagle_enabled = True
             else:
+                speculative_model = self.speculative_config.get("model")
+                if speculative_model in ("ngram", "[ngram]"):
+                    is_ngram_enabled = True
+            if not (is_ngram_enabled or is_eagle_enabled):
+                # Other speculative decoding methods are not supported yet.
                 _raise_or_fallback(feature_name="Speculative Decoding",
                                    recommend_to_remove=False)
                 return False
@@ -1523,6 +1529,10 @@ class EngineArgs:
         if is_ngram_enabled and _warn_or_fallback("ngram"):
             return False
 
+        # Eagle is under development, so we don't support it yet.
+        if is_eagle_enabled and _warn_or_fallback("Eagle"):
+            return False
+
         # Non-CUDA is supported on V1, but off by default for now.
         not_cuda = not current_platform.is_cuda()
         if not_cuda and _warn_or_fallback(  # noqa: SIM103

vllm/v1/spec_decode/eagle.py

+# SPDX-License-Identifier: Apache-2.0
+import torch
+import torch.nn as nn
+import triton
+import triton.language as tl
+
+from vllm.config import VllmConfig
+from vllm.forward_context import set_forward_context
+from vllm.v1.attention.backends.flash_attn import FlashAttentionMetadata
+from vllm.v1.sample.metadata import SamplingMetadata
+
+
+class EagleProposer:
+
+    def __init__(
+        self,
+        vllm_config: VllmConfig,
+        device: torch.device,
+    ):
+        self.vllm_config = vllm_config
+        self.num_speculative_tokens = (
+            vllm_config.speculative_config.num_speculative_tokens)
+        self.block_size = vllm_config.cache_config.block_size
+        self.arange = torch.arange(vllm_config.scheduler_config.max_num_seqs,
+                                   device=device)
+
+    def propose(
+        self,
+        # [num_tokens]
+        target_token_ids: torch.Tensor,
+        # [num_tokens]
+        target_positions: torch.Tensor,
+        # [num_tokens, hidden_size]
+        target_hidden_states: torch.Tensor,
+        # [num_tokens]
+        target_slot_mapping: torch.Tensor,
+        # [batch_size]
+        next_token_ids: torch.Tensor,
+        # [batch_size + 1] starting with 0
+        cu_num_tokens: torch.Tensor,
+        # [batch_size, max_num_blocks_per_req]
+        block_table: torch.Tensor,
+        sampling_metadata: SamplingMetadata,
+    ) -> tuple[torch.Tensor, torch.Tensor]:
+        num_tokens = target_token_ids.shape[0]
+        batch_size = next_token_ids.shape[0]
+        last_token_indices = cu_num_tokens[1:] - 1
+
+        input_ids = torch.empty_like(target_token_ids)
+        # Shift the input ids by one token.
+        # E.g., [a1, b1, b2, c1, c2, c3] -> [b1, b2, c1, c2, c3, c3]
+        input_ids[:-1] = target_token_ids[1:]
+        # Replace the last token with the next token.
+        # E.g., [b1, b2, c1, c2, c3, c3] -> [a2, b2, b3, c2, c3, c4]
+        input_ids[last_token_indices] = next_token_ids
+
+        seq_lens = target_positions[last_token_indices] + 1
+        # FIXME(woosuk): The below two ops cause synchronization. Optimize.
+        max_seq_len = seq_lens.max().item()
+        max_num_tokens = (cu_num_tokens[1:] - cu_num_tokens[:-1]).max().item()
+        attn_metadata = FlashAttentionMetadata(
+            num_actual_tokens=num_tokens,
+            max_query_len=max_num_tokens,
+            query_start_loc=cu_num_tokens,
+            max_seq_len=max_seq_len,
+            seq_lens=seq_lens,
+            block_table=block_table,
+            slot_mapping=target_slot_mapping,
+            # TODO(woosuk): Support cascade attention.
+            use_cascade=False,
+            common_prefix_len=0,
+            cu_prefix_query_lens=None,
+            prefix_kv_lens=None,
+            suffix_kv_lens=None,
+        )
+
+        with set_forward_context(attn_metadata, self.vllm_config):
+            hidden_states = self.model(
+                input_ids=input_ids,
+                hidden_states=target_hidden_states,
+                positions=target_positions,
+            )
+        sample_hidden_states = hidden_states[last_token_indices]
+        logits = self.model.compute_logits(sample_hidden_states, None)
+        draft_token_ids, draft_probs = compute_probs_and_sample_next_token(
+            logits, sampling_metadata)
+
+        # Early exit if there is only one draft token to be generated.
+        if self.num_speculative_tokens == 1:
+            # [batch_size, 1] and [batch_size, 1, vocab_size]
+            return draft_token_ids.view(-1, 1), draft_probs.unsqueeze(dim=1)
+
+        # Generate the remaining draft tokens.
+        draft_token_ids_list = [draft_token_ids]
+        draft_probs_list = [draft_probs]
+
+        positions = target_positions[last_token_indices]
+        hidden_states = sample_hidden_states
+        attn_metadata.num_actual_tokens = batch_size
+        attn_metadata.max_query_len = 1
+        attn_metadata.query_start_loc = self.arange[:batch_size]
+        for _ in range(self.num_speculative_tokens - 1):
+            # Update the inputs.
+            input_ids = draft_token_ids_list[-1]
+            positions += 1
+            attn_metadata.max_seq_len += 1
+            attn_metadata.seq_lens += 1
+            # Compute the slot mapping.
+            block_numbers = positions // self.block_size
+            block_ids = block_table.gather(dim=1,
+                                           index=block_numbers.view(-1, 1))
+            block_ids = block_ids.view(-1)
+            attn_metadata.slot_mapping = (block_ids * self.block_size +
+                                          positions % self.block_size)
+
+            # Run the model.
+            with set_forward_context(attn_metadata, self.vllm_config):
+                hidden_states = self.model(
+                    input_ids=input_ids,
+                    hidden_states=hidden_states,
+                    positions=positions,
+                )
+            logits = self.model.compute_logits(hidden_states, None)
+            draft_token_ids, probs = compute_probs_and_sample_next_token(
+                logits, sampling_metadata)
+            draft_token_ids_list.append(draft_token_ids)
+            draft_probs_list.append(probs)
+
+        # [batch_size, num_speculative_tokens]
+        draft_token_ids = torch.stack(draft_token_ids_list, dim=1)
+        # [batch_size, num_speculative_tokens, vocab_size]
+        draft_probs = torch.stack(draft_probs_list, dim=1)
+        return draft_token_ids, draft_probs
+
+    @staticmethod
+    def prepare_inputs(
+        # [batch_size + 1]
+        cu_target_query_lens: torch.Tensor,
+        # [batch_size]
+        num_rejected_tokens: torch.Tensor,
+    ) -> tuple[torch.Tensor, torch.Tensor]:
+        # cu_target_query_lens: [0, a, a + b, a + b + c]
+        # num_rejected_tokens: [n1, n2, n3]
+        # num_tokens_per_req: [a - n1, b - n2, c - n3]
+        # cu_num_tokens: [0, a - n1, a + b - n1 - n2, a + b + c - n1 - n2 - n3]
+        # token_indices: [0, 1, ..., a - n1 - 1,
+        #                 a, a + 1, ..., a + b - n2 - 1,
+        #                 a + b, a + b + 1, ..., a + b + c - n3 - 1]
+
+        # [0, a, a + b, a + b + c] -> [a, b, c]
+        query_len_per_req = (cu_target_query_lens[1:] -
+                             cu_target_query_lens[:-1])
+        # [a, b, c] -> [a - n1, b - n2, c - n3]
+        num_tokens_per_req = query_len_per_req - num_rejected_tokens
+
+        cu_num_tokens = torch.empty_like(cu_target_query_lens)
+        torch.cumsum(num_tokens_per_req, dim=0, out=cu_num_tokens[1:])
+        cu_num_tokens[0] = 0
+
+        # FIXME(woosuk): Avoid synchronization.
+        num_tokens = cu_num_tokens[-1].item()
+        token_indices = torch.empty(
+            num_tokens,
+            dtype=torch.int32,
+            device=cu_num_tokens.device,
+        )
+
+        batch_size = num_rejected_tokens.shape[0]
+        BLOCK_SIZE = 1024
+        prepare_input_kernel[(batch_size, )](
+            token_indices,
+            cu_target_query_lens,
+            cu_num_tokens,
+            BLOCK_SIZE=BLOCK_SIZE,
+        )
+        return cu_num_tokens, token_indices
+
+    def load_model(self, target_model: nn.Module) -> None:
+        self.model = DummyEagleModel()
+        self.model.get_input_embeddings = target_model.get_input_embeddings
+        self.model.compute_logits = target_model.compute_logits
+
+
+# FIXME(woosuk): This is a dummy model for testing.
+# Remove this once we have a real model.
+class DummyEagleModel(nn.Module):
+
+    def __init__(self):
+        super().__init__()
+
+    def forward(
+        self,
+        input_ids: torch.Tensor,
+        hidden_states: torch.Tensor,
+        positions: torch.Tensor,
+    ) -> torch.Tensor:
+        input_embeddings = self.get_input_embeddings(input_ids)
+        return hidden_states + input_embeddings  # Dummy return.
+
+
+# FIXME(woosuk): The logic here is duplicated with the main sampling code.
+# We should refactor this to reuse the same sampling implementation.
+def compute_probs_and_sample_next_token(
+    logits: torch.Tensor,
+    sampling_metadata: SamplingMetadata,
+) -> tuple[torch.Tensor, torch.Tensor]:
+    if sampling_metadata.all_greedy:
+        # For greedy requests, draft_probs is not used in rejection sampling.
+        # Therefore, we can just return the logits.
+        probs = logits
+        next_token_ids = logits.argmax(dim=-1)
+        return next_token_ids, probs
+
+    is_greedy = sampling_metadata.temperature == -1
+    temperature = torch.where(is_greedy, 1.0, sampling_metadata.temperature)
+    logits.div_(temperature.view(-1, 1))
+    probs = logits.softmax(dim=-1, dtype=torch.float32)
+
+    # NOTE(woosuk): Currently, we ignore most of the sampling parameters in
+    # generating the draft tokens. We only use the temperature. While this
+    # could degrade the acceptance rate, it does not affect the distribution
+    # of the generated tokens after rejection sampling.
+
+    # TODO(woosuk): Consider seeds.
+    q = torch.empty_like(probs)
+    q.exponential_()
+    next_token_ids = probs.div_(q).argmax(dim=-1).view(-1)
+    if not sampling_metadata.all_random:
+        greedy_token_ids = probs.argmax(dim=-1)
+        next_token_ids = torch.where(
+            is_greedy,
+            greedy_token_ids,
+            next_token_ids,
+        )
+    return next_token_ids, probs
+
+
+@triton.jit
+def prepare_input_kernel(
+    out_ptr,
+    cu_query_lens_ptr,
+    cu_num_tokens_ptr,
+    BLOCK_SIZE: tl.constexpr,
+):
+    pid = tl.program_id(0)
+
+    # [start_pos, end_pos)
+    start_pos = tl.load(cu_num_tokens_ptr + pid)
+    end_pos = tl.load(cu_num_tokens_ptr + pid + 1)
+    num_tokens = end_pos - start_pos
+
+    index_start = tl.load(cu_query_lens_ptr + pid)
+    indices = index_start + tl.arange(0, BLOCK_SIZE)
+
+    num_blocks = tl.cdiv(num_tokens, BLOCK_SIZE)
+    for i in tl.range(num_blocks):
+        offset = i * BLOCK_SIZE + tl.arange(0, BLOCK_SIZE)
+        tl.store(
+            out_ptr + start_pos + offset,
+            indices,
+            mask=offset < num_tokens,
+        )

vllm/v1/spec_decode/ngram_proposer.py

@@ -4,9 +4,14 @@ from typing import Optional
 import numpy as np
 from numba import jit
 
+from vllm.config import VllmConfig
+
 
 class NgramProposer:
 
+    def __init__(self, vllm_config: VllmConfig):
+        self.vllm_config = vllm_config
+
     def propose(
         self,
         context_token_ids: np.ndarray,
@@ -50,6 +55,10 @@ class NgramProposer:
                 return result
         return None
 
+    def load_model(self, *args, **kwargs):
+        # No model to load.
+        pass
+
 
 @jit(nopython=True)
 def _kmp_lps_array(pattern: np.ndarray) -> np.ndarray:

vllm/v1/worker/gpu_input_batch.py

@@ -39,9 +39,18 @@ class CachedRequestState:
 
     lora_request: Optional[LoRARequest] = None
 
+    def __post_init__(self):
+        self.num_prompt_tokens = len(self.prompt_token_ids)
+
     @property
     def num_tokens(self) -> int:
-        return len(self.prompt_token_ids) + len(self.output_token_ids)
+        return self.num_prompt_tokens + len(self.output_token_ids)
+
+    def get_token_id(self, idx: int) -> int:
+        if idx < self.num_prompt_tokens:
+            return self.prompt_token_ids[idx]
+        else:
+            return self.output_token_ids[idx - self.num_prompt_tokens]
 
 
 class InputBatch:

vllm/v1/worker/gpu_model_runner.py

@@ -35,6 +35,7 @@ from vllm.v1.outputs import (EMPTY_MODEL_RUNNER_OUTPUT, LogprobsTensors,
                              ModelRunnerOutput)
 from vllm.v1.sample.metadata import SamplingMetadata
 from vllm.v1.sample.rejection_sampler import RejectionSampler
+from vllm.v1.spec_decode.eagle import EagleProposer
 from vllm.v1.spec_decode.metadata import SpecDecodeMetadata
 from vllm.v1.spec_decode.ngram_proposer import NgramProposer
 from vllm.v1.spec_decode.utils import is_spec_decode_supported
@@ -157,18 +158,15 @@ class GPUModelRunner(LoRAModelRunnerMixin):
         self.use_spec_decode = False
         if self.speculative_config:
             self.use_spec_decode = True
-            assert self.speculative_config.method == "ngram", \
-                    "Currently, only ngram spec decode is supported in V1."
             if get_pp_group().is_last_rank:
-                self.drafter = NgramProposer()
-                # Trigger Numba JIT compilation for N-gram proposer.
-                # This usually takes less than 1 second.
-                self.drafter.propose(
-                    np.zeros(1024, dtype=np.int32),
-                    self.speculative_config.prompt_lookup_min,
-                    self.speculative_config.prompt_lookup_max,
-                    self.speculative_config.num_speculative_tokens,
-                )
+                if self.speculative_config.method == "ngram":
+                    self.drafter = NgramProposer(self.vllm_config)
+                elif self.speculative_config.method == "eagle":
+                    self.drafter = EagleProposer(self.vllm_config,
+                                                 self.device)  # type: ignore
+                else:
+                    raise ValueError("Unknown speculative decoding method: "
+                                     f"{self.speculative_config.method}")
                 self.rejection_sampler = RejectionSampler()
 
         # Request states.
@@ -1144,10 +1142,75 @@ class GPUModelRunner(LoRAModelRunnerMixin):
             valid_sampled_token_ids[i].clear()
 
         if not self.use_spec_decode:
+            # Speculative decoding is not enabled.
             spec_token_ids = None
-        else:
+        elif self.speculative_config.method == "ngram":
+            assert isinstance(self.drafter, NgramProposer)
             spec_token_ids = self.generate_draft_token_ids(
                 valid_sampled_token_ids, sampling_metadata)
+        elif self.speculative_config.method == "eagle":
+            assert isinstance(self.drafter, EagleProposer)
+            # TODO(woosuk): Refactor the loop.
+            next_token_ids: list[int] = []
+            for i, token_ids in enumerate(valid_sampled_token_ids):
+                if token_ids:
+                    # Common case.
+                    next_token_id = token_ids[-1]
+                else:
+                    # Partial prefill (rare case).
+                    # Get the next token id from the request state.
+                    req_id = self.input_batch.req_ids[i]
+                    req_state = self.requests[req_id]
+                    seq_len = (req_state.num_computed_tokens +
+                               scheduler_output.num_scheduled_tokens[req_id])
+                    next_token_id = req_state.get_token_id(seq_len)
+                next_token_ids.append(next_token_id)
+            next_token_ids = torch.tensor(next_token_ids,
+                                          dtype=torch.int32,
+                                          device=self.device)
+
+            if spec_decode_metadata is None:
+                # input_ids can be None for multimodal models.
+                target_token_ids = self.input_ids[:num_scheduled_tokens]
+                target_positions = positions
+                target_hidden_states = hidden_states
+                target_slot_mapping = attn_metadata.slot_mapping
+                cu_num_tokens = attn_metadata.query_start_loc
+            else:
+                # TODO(woosuk): Refactor this.
+                num_draft_tokens = spec_decode_metadata.num_draft_tokens
+                num_rejected_tokens = [
+                    n + 1 - len(valid_sampled_token_ids[i]) if n > 0 else 0
+                    for i, n in enumerate(num_draft_tokens)
+                ]
+                num_rejected_tokens = torch.tensor(
+                    num_rejected_tokens,
+                    dtype=torch.int32,
+                    device=self.device,
+                )
+                cu_num_tokens, token_indices = self.drafter.prepare_inputs(
+                    attn_metadata.query_start_loc,
+                    num_rejected_tokens,
+                )
+                target_token_ids = self.input_ids[token_indices]
+                target_positions = positions[token_indices]
+                target_hidden_states = hidden_states[token_indices]
+                target_slot_mapping = attn_metadata.slot_mapping[token_indices]
+
+            draft_token_ids, draft_probs = self.drafter.propose(
+                target_token_ids=target_token_ids,
+                target_positions=target_positions,
+                target_hidden_states=target_hidden_states,
+                target_slot_mapping=target_slot_mapping,
+                next_token_ids=next_token_ids,
+                cu_num_tokens=cu_num_tokens,
+                block_table=attn_metadata.block_table,
+                sampling_metadata=sampling_metadata,
+            )
+            spec_token_ids = draft_token_ids.tolist()
+            # TODO(woosuk): Cache draft_probs and use it for rejection sampling
+            # in the next step.
+            del draft_probs
 
         return ModelRunnerOutput(
             req_ids=self.input_batch.req_ids,
@@ -1205,6 +1268,9 @@ class GPUModelRunner(LoRAModelRunnerMixin):
                                                   self.scheduler_config,
                                                   self.lora_config,
                                                   self.device)
+            if hasattr(self, "drafter"):
+                logger.info("Loading drafter model...")
+                self.drafter.load_model(self.model)
             time_after_load = time.perf_counter()
         self.model_memory_usage = m.consumed_memory
         logger.info("Model loading took %.4f GiB and %.6f seconds",