tpu-inference 0.11.1.dev202511270815__py3-none-any.whl → 0.13.0rc2.post7__py3-none-any.whl

tpu_inference/runner/kv_cache_manager.py

@@ -1,3 +1,17 @@
+# Copyright 2025 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
 import functools
 from typing import TYPE_CHECKING, Dict, List
 
@@ -39,20 +53,30 @@ class KVCacheManager:
         # means this layer will perform attention using the keys and values
         # from the KV cache of `shared_kv_cache_layers[layer_name]`.
         self.shared_kv_cache_layers: dict[str, str] = {}
+        self.use_mla = self.runner.model_config.use_mla
 
     def get_kv_cache_spec(self):
         # TODO(xiang): this hack tricks engine core to init successfully
         block_size = self.runner.cache_config.block_size
-        use_mla = self.runner.model_config.use_mla
         kv_cache_spec: dict[str, KVCacheSpec] = {}
 
         # If use pure jax (MODEL_IMPL_TYPE=flax_nnx), we don't register
         # attention into compilation config.
         # Use FullAttentionSpec for each layer
         # TODO(pooyam): Is it possible to merge the logic for vllm and non-vllm models?
+        model_config = self.runner.model_config
+        if self.use_mla:
+            # Individually pad the RopE and latents
+            qk_rope_head_dim = getattr(model_config.hf_text_config,
+                                       "qk_rope_head_dim", 0)
+            padded_kv_lora_rank = common_utils.align_to(
+                model_config.hf_text_config.kv_lora_rank, 128)
+            padded_qk_rope_head_dim = common_utils.align_to(
+                qk_rope_head_dim, 128)
+            mla_head_size = padded_kv_lora_rank + padded_qk_rope_head_dim
+
         if len(self.runner.vllm_config.compilation_config.
                static_forward_context) == 0:
-            model_config = self.runner.model_config
             parallel_config = self.runner.parallel_config
             # Pad num_kv_heads to multiple of TP size.
             num_kv_heads = common_utils.get_padded_num_heads(
@@ -61,11 +85,11 @@ class KVCacheManager:
             head_size = common_utils.get_padded_head_dim(
                 model_config.get_head_size())
             for i in range(model_config.get_num_layers(parallel_config)):
-                if use_mla:
+                if self.use_mla:
                     kv_cache_spec[f"layer.{i}"] = MLAAttentionSpec(
                         block_size=block_size,
-                        num_kv_heads=num_kv_heads,
-                        head_size=head_size,
+                        num_kv_heads=1,
+                        head_size=mla_head_size,
                         dtype=self.runner.kv_cache_dtype,
                         cache_dtype_str=self.runner.vllm_config.cache_config.
                         cache_dtype)
@@ -83,14 +107,13 @@ class KVCacheManager:
                     self.runner.mesh.shape["model"])
                 head_size = common_utils.get_padded_head_dim(
                     hf_config.hidden_size // hf_config.num_attention_heads)
-
                 # Eagle3 has only 1 layer
                 for i in range(1):
-                    if use_mla:
-                        kv_cache_spec[f"layer.{i}"] = MLAAttentionSpec(
+                    if self.use_mla:
+                        kv_cache_spec[f"draft_layer.{i}"] = MLAAttentionSpec(
                             block_size=block_size,
-                            num_kv_heads=num_kv_heads,
-                            head_size=head_size,
+                            num_kv_heads=1,
+                            head_size=mla_head_size,
                             dtype=self.runner.kv_cache_dtype,
                             cache_dtype_str=self.runner.vllm_config.
                             cache_config.cache_dtype)
@@ -104,6 +127,7 @@ class KVCacheManager:
             # Else propagate attention modules from compilation config.
             layers = get_layers_from_vllm_config(self.runner.vllm_config,
                                                  Attention)
+            logger.warning(f"Compilation num_layers = {len(layers.items())}")
             for layer_name, attn_module in layers.items():
                 if (kv_tgt_layer :=
                         attn_module.kv_sharing_target_layer_name) is not None:
@@ -127,11 +151,11 @@ class KVCacheManager:
                                 attn_module.head_size),
                             dtype=self.runner.kv_cache_dtype,
                             sliding_window=attn_module.sliding_window)
-                    elif use_mla:
-                        kv_cache_spec[f"layer.{i}"] = MLAAttentionSpec(
+                    elif self.use_mla:
+                        kv_cache_spec[layer_name] = MLAAttentionSpec(
                             block_size=block_size,
-                            num_kv_heads=attn_module.num_kv_heads,
-                            head_size=attn_module.head_size,
+                            num_kv_heads=1,
+                            head_size=mla_head_size,
                             dtype=self.runner.kv_cache_dtype,
                             cache_dtype_str=self.runner.vllm_config.
                             cache_config.cache_dtype)
@@ -198,14 +222,20 @@ class KVCacheManager:
             # num_blocks must be a multiple of dp_size
             num_blocks = (num_blocks // dp_size) * dp_size
             # NOTE: we'll multiply the num_kv_heads by 2 in the function
+            if self.use_mla:
+                head_size = self.runner.model_config.hf_config.kv_lora_rank + \
+                    self.runner.model_config.hf_config.qk_rope_head_dim
+            else:
+                head_size = representative_spec.head_size
             kv_cache = create_kv_caches(
                 num_blocks=num_blocks,
                 block_size=representative_spec.block_size,
                 num_kv_heads=representative_spec.num_kv_heads,
-                head_size=representative_spec.head_size,
+                head_size=head_size,
                 mesh=self.runner.mesh,
                 layer_names=[f'kv_cache_tensor.{i}'],
                 cache_dtype=t2j_dtype(representative_spec.dtype),
+                use_mla=self.use_mla,
             )[0]
             kv_caches.append(kv_cache)
             num_blocks_list.append(num_blocks)
@@ -289,13 +319,8 @@ class KVCacheManager:
 
         def _update_layer(cache, slices):
             """The function to apply to each layer's cache and slices."""
-            reshaped_slices = slices.reshape(-1, 1, block_size,
-                                             *slices.shape[1:])
-            for (i, block_idx) in enumerate(block_numbers):
-                cache = jax.lax.dynamic_update_slice_in_dim(cache,
-                                                            reshaped_slices[i],
-                                                            block_idx,
-                                                            axis=0)
+            reshaped_slices = slices.reshape(-1, block_size, *slices.shape[1:])
+            cache.at[block_numbers].set(reshaped_slices)
             return cache
 
         return jax.tree.map(_update_layer, kv_caches, kv_cache_slices)
@@ -348,16 +373,12 @@ class KVCacheManager:
         """
         if block_ids == list(range(block_ids[0],
                                    block_ids[0] + len(block_ids))):
-            with runner_utils.LatencyTracker(
-                    "BatchedGatherKVSlices-for-blocks"):
-                batched_kv_cache_per_layer = self._jitted_gather_continuous_kv_cache(
-                    self.runner.kv_caches, block_ids[0], len(block_ids))
+            batched_kv_cache_per_layer = self._jitted_gather_continuous_kv_cache(
+                self.runner.kv_caches, block_ids[0], len(block_ids))
 
         else:
-            with runner_utils.LatencyTracker(
-                    "BatchedGatherKVSlices-for-blocks"):
-                batched_kv_cache_per_layer = self._jitted_gather_kv_cache(
-                    self.runner.kv_caches, jnp.array(block_ids))
+            batched_kv_cache_per_layer = self._jitted_gather_kv_cache(
+                self.runner.kv_caches, jnp.array(block_ids))
         return batched_kv_cache_per_layer
 
     def transfer_kv_cache(self,
@@ -446,6 +467,7 @@ class KVCacheManager:
                     kv_cache_slices,
                     start_block,
                 )
+                jax.block_until_ready(self.runner.kv_caches)
         else:
             with runner_utils.LatencyTracker(
                     f"JittedInsertKVCache-b{len(block_numbers)}"):
@@ -457,6 +479,7 @@ class KVCacheManager:
                     kv_cache_slices,
                     jnp.array(block_numbers),
                 )
+                jax.block_until_ready(self.runner.kv_caches)
 
         logger.debug(
             f"Updated kv cache entries cnt={len(self.runner.kv_caches)}")

tpu_inference/runner/lora_utils.py

@@ -1,3 +1,17 @@
+# Copyright 2025 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
 from __future__ import annotations
 
 from typing import TYPE_CHECKING

tpu_inference/runner/multimodal_manager.py

@@ -1,3 +1,17 @@
+# Copyright 2025 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
 from typing import TYPE_CHECKING
 
 import jax
@@ -134,7 +148,7 @@ class MultiModalManager:
             # 2. A list or tuple (length: num_items) of tensors, each of shape
             # (feature_size, hidden_size) in case the feature size is dynamic
             # depending on the input multimodal items.
-            curr_group_outputs = self.runner.get_multimodal_embeddings_fn(
+            curr_group_outputs = self.runner.embed_multimodal_fn(
                 self.runner.state, image_grid_thw, **batched_mm_inputs)
 
             sanity_check_mm_encoder_outputs(

tpu_inference/runner/persistent_batch_manager.py

@@ -1,3 +1,17 @@
+# Copyright 2025 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
 from typing import Dict
 
 import jax
@@ -14,12 +28,13 @@ class PersistentBatchManager:
     def __init__(self, requests: Dict[str, CachedRequestState],
                  input_batch: InputBatch, encoder_cache: Dict[str,
                                                               'jax.Array'],
-                 uses_mrope: bool, model_config):
+                 uses_mrope: bool, model_config, is_last_rank: bool):
         self.requests = requests
         self.input_batch = input_batch
         self.encoder_cache = encoder_cache
         self.uses_mrope = uses_mrope
         self.model_config = model_config
+        self.is_last_rank = is_last_rank
 
     def _reorder_batch(self, scheduler_output: "VllmSchedulerOutput") -> int:
         """ Reorder the sheduled requests to RPA kernel friendly distribution
@@ -179,9 +194,35 @@ class PersistentBatchManager:
             num_computed_tokens = req_data.num_computed_tokens[i]
             new_block_ids = req_data.new_block_ids[i]
             resumed_from_preemption = req_data.resumed_from_preemption[i]
+            num_output_tokens = req_data.num_output_tokens[i]
 
             # Update the cached states.
             req_state.num_computed_tokens = num_computed_tokens
+            req_index = self.input_batch.req_id_to_index.get(req_id)
+
+            if not self.is_last_rank:
+                # When using PP, the scheduler sends the sampled tokens back,
+                # because there's no direct communication between the first-
+                # stage worker and the last-stage worker.
+                new_token_ids = req_data.new_token_ids[i]
+                # Add the sampled token(s) from the previous step (if any).
+                # This doesn't include "unverified" tokens like spec tokens.
+                num_new_tokens = (num_computed_tokens + len(new_token_ids) -
+                                  req_state.num_tokens)
+                if num_new_tokens == 1:
+                    req_state.output_token_ids.append(new_token_ids[-1])
+                elif num_new_tokens > 0:
+                    req_state.output_token_ids.extend(
+                        new_token_ids[-num_new_tokens:])
+            elif num_output_tokens < len(req_state.output_token_ids):
+                del req_state.output_token_ids[num_output_tokens:]
+                if req_index is not None:
+                    end_idx = (self.input_batch.num_prompt_tokens[req_index] +
+                               num_output_tokens)
+                    self.input_batch.num_tokens[req_index] = end_idx
+                    self.input_batch.num_tokens_no_spec[req_index] = end_idx
+
+            # Update the block IDs.
             if not resumed_from_preemption:
                 if new_block_ids is not None:
                     # Append the new blocks to the existing block IDs.
@@ -194,7 +235,6 @@ class PersistentBatchManager:
                 # Replace the existing block IDs with the new ones.
                 req_state.block_ids = new_block_ids
 
-            req_index = self.input_batch.req_id_to_index.get(req_id)
             if req_index is None:
                 # The request is not in the persistent batch.
                 # The request was either preempted and resumed later, or was not
@@ -209,6 +249,18 @@ class PersistentBatchManager:
                 self.input_batch.block_table.append_row(
                     new_block_ids, req_index)
 
+            # For the last rank, we don't need to update the token_ids_cpu
+            # because the sampled tokens are already cached.
+            if not self.is_last_rank:
+                start_token_index = num_computed_tokens
+                end_token_index = num_computed_tokens + len(new_token_ids)
+                self.input_batch.token_ids_cpu[
+                    req_index,
+                    start_token_index:end_token_index] = new_token_ids
+                self.input_batch.num_tokens_no_spec[
+                    req_index] = end_token_index
+                self.input_batch.num_tokens[req_index] = end_token_index
+
             # Add spec_token_ids to token_ids_cpu.
             spec_token_ids = scheduler_output.scheduled_spec_decode_tokens.get(
                 req_id, ())

tpu_inference/runner/speculative_decoding_manager.py

@@ -1,3 +1,17 @@
+# Copyright 2025 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
 from __future__ import annotations
 
 from dataclasses import dataclass

tpu_inference/runner/structured_decoding_manager.py

@@ -1,3 +1,17 @@
+# Copyright 2025 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
 import functools
 from typing import TYPE_CHECKING, Tuple