apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2026-04-25 09:57:51 +08:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity
5,093 Commits 40 Branches 17 Tags
d8cdda86cb022f210fdeafadaee054e26f7acf87
Commit Graph

392 Commits

Author SHA1 Message Date
kevin 7707be8384 [Feature][KVCache] Implement Cache Manager V1 with GPU + CPU Cache Support (1/n) (#7097) 
* [Feature][KVCache] Support cache manager v1 architecture

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* Update cache manager and related modules

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* chore: update cache_manager and related modules

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: add node to evictable set in complete_swap_to_device

When a node transitions from SWAP_TO_DEVICE to DEVICE via
complete_swap_to_device, it was not being added to the
_evictable_device set. This caused nodes with ref_count=0 to
become "orphaned" - not appearing in any evictable set despite
having cache_status=DEVICE.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat: update cache manager v1 and related modules

- Add new cache_manager.py with cache management functionality
- Add radix_tree.py for prefix caching
- Update block_pool.py and metadata.py
- Update request.py and resource_manager_v1.py for scheduling
- Update gpu_model_runner.py for GPU model execution

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* feat(cache): add cache controller v1 implementation

- Add CacheController class for cache management
- Update config.py with cache related configurations
- Refactor gpu_model_runner.py for improved cache handling

* feat(cache_manager): update cache manager v1

* fix(cache_manager): 修复 swap_cache H2D/D2H 方向的 block_ids 逻辑并清理 ForwardMeta

## Motivation

修复 swap_cache_optimized.cu 中 H2D 方向时 src/dst block_ids 使用错误的问题，
并清理 ForwardMeta 中已废弃的 cache_controller 字段。

## Modifications

- fix: swap_cache_optimized.cu 中根据 D2H 模板参数正确选取 src/dst block_ids，
  修复 H2D 方向 src/dst 倒置 bug（同时修复 SwapCachePerLayerImpl 和 SwapCacheAllLayersBatchImpl）
- refactor: cache_manager/v1/__init__.py 将 LayerSwapTimeoutError 导入从
  cache_controller 改为 cache_utils（正确来源）
- refactor: ForwardMeta 移除废弃的 cache_controller 字段
- refactor: gpu_model_runner.py 移除对应的 cache_controller 赋值语句
- test: 新增 tests/cache_manager/v1/test_swap_cache_ops.py 单元测试

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* feat(cache_manager): refactor cache manager v1 and optimize swap ops

## Motivation

对 cache manager v1 进行重构和优化，精简代码结构，提升可维护性。

## Modifications

- 重构 transfer_manager.py，大幅精简代码逻辑
- 优化 swap_cache_optimized.cu GPU 算子实现
- 调整 cache_manager.py、cache_controller.py 逻辑，修复 free_device_blocks 方法缺失问题
- 更新 block_pool.py、cache_utils.py、metadata.py、radix_tree.py
- 精简 gpu_model_runner.py、forward_meta.py、attention.py 中相关调用
- 更新对应单元测试（test_cache_controller、test_swap_cache_ops、test_transfer_manager）
- 调整 config.py 中相关配置项

* [KVCache][MTP] 支持 cache_manager_v1 下的 MTP KV Cache 初始化及多模态 hash

## Motivation

在 enable_cache_manager_v1 路径下，MTP（speculative decode）的 KV Cache 需要由
CacheController 统一管理，以复用 swap/transfer 能力，同时修复多模态场景下 block
hash 未携带 multimodal extra_keys 的问题。

## Modifications

- `cache_controller.py`
  - 新增 `initialize_mtp_kv_cache`：通过 CacheController 初始化 MTP KV Cache，
    并将其注册到 cache_kvs_map，使 transfer_manager 自动覆盖 MTP 层
  - `initialize_host_cache` 中的 num_layers 改为包含 MTP 额外 cache 层数，保证
    Host Cache 也为 MTP 分配足够空间
  - `_free_gpu_cache` 改名为 `free_gpu_cache`（对外可调用）

- `cache_utils.py`
  - 新增 `get_block_hash_extra_keys`：提取单个 block 内的多模态 hash 信息，
    对齐 PrefixCacheManager 的 multimodal extra_keys 逻辑
  - `get_request_block_hasher` 中在 hash_block_tokens 时携带 extra_keys，
    修复多模态场景 prefix cache 命中率不准的问题

- `spec_decode/mtp.py`
  - `update_mtp_block_num` 新增 `skip_cache_init` 参数，避免 v1 cache manager
    路径下重复初始化 MTP KV Cache

- `gpu_model_runner.py`
  - `initialize_kv_cache(v1)` 路径：在主模型 cache 初始化后，调用
    `cache_controller.initialize_mtp_kv_cache` 完成 MTP cache 创建
  - `clear_cache` / `wakeup` / `reset` 等路径：respect `enable_cache_manager_v1`
    标志，跳过重复的 proposer.initialize_kv_cache 调用

## Usage or Command

```bash
# 启动支持 MTP + cache_manager_v1 的推理服务（示例）
bash run.sh
```

* fix(cache_manager): multi-GPU fix, mm hash boundary fix, and remove batch ops

1. Fix CuPy stream/event creation for multi-GPU: wrap all stream operations
   with cp.cuda.Device(device_id) context to ensure streams/events are bound
   to the correct device, preventing cross-device errors in multi-GPU setups.

2. Remove cudaSetDevice from SwapCacheAllLayers (handled by cupy context now).

3. Remove swap_cache_all_layers_batch op: simplified the implementation by
   removing the batch upload variant; all-layer transfers now use the standard
   swap_cache_all_layers with cupy device context.

4. Fix mm hash boundary comparison in get_block_hash_extra_keys: change
   strict less-than (<) to less-than-or-equal (<=) so that multimodal items
   ending exactly at block start are correctly excluded.

5. Extract config fields to KVCacheBase: model_config, cache_config,
   quant_config, parallel_config are now set in the base class __init__ to
   avoid duplication in CacheController and CacheManager subclasses.

6. Translate metadata.py docstrings from Chinese to English for broader
   contributor accessibility.

7. Add test_cache_utils.py: comprehensive unit tests for
   get_block_hash_extra_keys covering all boundary and overlap scenarios.

8. Expand test suite: test_request.py cache fields tests, test_radix_tree.py
   backup candidate tests, test_transfer_manager.py and test_cache_manager.py
   multi-GPU and concurrent operation tests.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] fix List import and move write_policy normalization to CacheManager

## Motivation

修复两处问题：
1. `fastdeploy/engine/request.py` 中 `List` 未导入导致 pre-commit F821 报错
2. `write_policy` 归一化逻辑（`write_through` → `write_through_selective`）不应放在 `FDConfig`，移至 `CacheManager.__init__` 中，使其只影响 Cache Manager V1 的内部逻辑

## Modifications

- `fastdeploy/engine/request.py`: 在 `typing` 导入中补充 `List`，删除重复的 `CacheSwapMetadata` TYPE_CHECKING 导入，修复 F821/F811
- `fastdeploy/config.py`: 删除 `write_policy` 归一化逻辑
- `fastdeploy/cache_manager/v1/cache_manager.py`: 将归一化逻辑移入 `CacheManager.__init__`

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] fix pre-commit code style issues

## Motivation

修复 CI pre-commit 代码风格检查失败问题。

## Modifications

- `fastdeploy/engine/common_engine.py`: black 格式化
- `fastdeploy/worker/worker_process.py`: black 格式化 + isort 修复
- `fastdeploy/cache_manager/v1/storage/__init__.py`: isort 修复
- `fastdeploy/worker/gpu_worker.py`: isort 修复

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [Feature][KVCache] update cache_manager_v1 modules

## Motivation

更新 Cache Manager V1 相关模块，完善版权信息、改进模块结构与可维护性。

## Modifications

- `fastdeploy/cache_manager/v1/` 系列模块：补充版权 header，优化代码结构
- `fastdeploy/config.py`：配置项更新
- `fastdeploy/engine/sched/resource_manager_v1.py`：调度相关更新

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [Feature][KVCache] add BatchRequest.from_tasks and refactor worker task parsing

## Motivation

将 worker_process 中重复的 task 解析逻辑收敛到 BatchRequest，减少代码冗余，提升可维护性。

## Modifications

- `fastdeploy/engine/request.py`：新增 `BatchRequest.from_tasks()` 类方法，统一将 task_queue 任务分类为推理请求和控制请求
- `fastdeploy/worker/worker_process.py`：使用 `BatchRequest.from_tasks()` 替代内联解析逻辑，并修复重复的 control_reqs 处理块

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [Feature][KVCache] add NUMA affinity for host cache and skip swap cache tests

## Motivation

优化 Host cache 内存分配的 NUMA 亲和性，减少跨 NUMA 访问延迟；
同时跳过 swap cache ops 测试（当前环境不支持）。

## Modifications

- `fastdeploy/cache_manager/v1/cache_controller.py`：
  - 新增 `_get_numa_node_for_gpu()` 方法，通过 nvidia-smi 或 sysfs 获取 GPU 对应的 NUMA 节点
  - 新增 `_bind_to_closest_numa_node()` 方法，绑定当前线程到 GPU 最近的 NUMA 节点
  - 在 `initialize_host_cache()` 中调用 NUMA 绑定，优化 H2D 传输性能
- `tests/cache_manager/v1/test_swap_cache_ops.py`：跳过所有测试类（`TestSwapCacheAllLayersCorrectness`、`TestSwapCacheAllLayersPerformance`、`TestSwapCacheRandomBlockIndices`）

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] fix unittest failures for cache_manager_v1

三个单测因接口变更或 Mock 方式问题导致失败，需修复。

- tests/distributed/chunked_moe.py：`setup_model_runner` 使用 `__new__` 跳过 `__init__`，补加 `enable_cache_manager_v1 = False`，修复 `AttributeError`
- tests/engine/test_resource_manager.py：`PrefixCacheManager` 为局部导入，`patch` 路径改为定义位置 `fastdeploy.cache_manager.prefix_cache_manager.PrefixCacheManager`
- tests/v1/test_resource_manager_v1.py：`_trigger_preempt` 第四参数已由 `list` 改为 `BatchRequest`，更新测试传参和断言

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] remove debug logging code

## Modifications

- fastdeploy/engine/request.py：删除调试用 logger 及 prompt_hashes 中的 debug 日志
- fastdeploy/worker/worker_process.py：删除 __main__ 中的调试 import 和 print 语句

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] fix cupy device id caching and pickle for _match_result

## Motivation

修复两个 bug：
1. `transfer_manager.py` 中每次调用 `cp.cuda.runtime.getDevice()` 存在隐患，应在初始化时缓存为实例变量，保证后续操作使用一致的设备 ID。
2. `request.py` 的 `__getstate__` 未跳过 `_match_result`，该字段包含 BlockNode 树的父子循环引用，pickle 时会触发 `RecursionError`；同时补充 `__setstate__` 确保 unpickle 后字段恢复为安全默认值。

## Modifications

- `transfer_manager.py`：初始化时调用 `cp.cuda.runtime.getDevice()` 并缓存到 `self._cupy_device_id`，后续 `with cp.cuda.Device(...)` 和日志均使用该缓存值。
- `request.py`：
  - `__getstate__` 中将 `_match_result` 加入跳过集合 `_SKIP_KEYS`，避免循环引用导致 pickle 失败。
  - 新增 `__setstate__`，unpickle 后将 `_block_hasher` 和 `_match_result` 恢复为 `None`。

## Usage or Command

* fix(test): fix unit test errors for _trigger_preempt and wakeup with MTP

- Use BatchRequest instead of list in test_trigger_preempt_records_tasks
- Add missing enable_cache_manager_v1 attr in TestSleepWakeupBehavior._make_runner

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] fix gpu_free_block_list returning wrong block IDs

## Motivation

`gpu_free_block_list` 的兼容 property 中误用了 `list(range(N))`，
将 `available_blocks()` 的返回值当作整数传给 `range()`，
导致返回 `[0, 1, ..., N-1]` 的假列表，而非真实的空闲 block ID。

## Modifications

- `cache_manager/v1/cache_manager.py`：将 `list(range(self._device_pool.available_blocks()))` 改为 `list(self._device_pool.available_blocks())`

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] 修复 gpu_free_block_list 返回 int 导致 TypeError

## Motivation

gpu_free_block_list 属性中调用 BlockPool.available_blocks()，
该方法返回 int（空闲块数量），用 list() 包装 int 会触发
TypeError: 'int' object is not iterable。

## Modifications

将 list(self._device_pool.available_blocks()) 改为
list(self._device_pool._free_blocks)，直接返回空闲块索引列表。

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [KVCache][CacheManager] 适配 V1 CacheManager 的 pause/sleep/free_cache 操作

## Motivation

V1 CacheManager 引入了新的 reset_cache() 接口，pause 和 sleep 操作需要适配，
同时 free_cache 需要支持可选的 clear_storage 参数。

## Modifications

- cache_controller.py: free_cache 新增 clear_storage 参数（默认 False），
  仅当 clear_storage=True 时才调用 _clear_storage()，避免不必要的 storage 清空
- common_engine.py: pause 和 sleep 操作中，当 ENABLE_V1_KVCACHE_MANAGER 时
  使用 cache_manager.reset_cache() 替代旧的 reset() 和 pause_transfer 逻辑
- gpu_model_runner.py: sleep 时仅在非 V1 cache manager 下执行 MTP cache 清除

## Usage or Command

# 启动服务（V1 CacheManager）
python -m fastdeploy.entrypoints.openai.api_server \
  --enable-v1-kvcache-manager \
  ...

* [BugFix][KVCache] fix missing enable_cache_manager_v1 in test mocks and remove unused select_blocks_for_backup

- Remove unused `select_blocks_for_backup` method from radix_tree.py
- Fix `match_prefix` default param `skip_storage=True` and log order in cache_manager.py
- Sync test_gpu_model_runner.py with upstream/develop (add TestInsertTasksV1SplitwiseSuffix)
- Add `enable_cache_manager_v1=False` to all mock runners to fix AttributeError in CI

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [BugFix][KVCache] simplify _free_blocks in ResourceManagerV1 for non-v1 path

Remove redundant prefix_caching branch in else path; always call
recycle_gpu_blocks with full block_tables for non-cache-manager-v1 case.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* [KVCache][Optimization][BugFix] fix and optimize block_pool, cache_manager, transfer_manager, request

## Motivation

修复 cache_manager v1 中若干代码质量问题，提升性能并消除潜在的类型不一致 Bug。

## Modifications

1. **block_pool.py**：`BlockPool.allocate` 将逐个 pop 循环替换为切片 + 批量 set.update，消除 Python 循环开销，O(n) → O(k)（C 层批量操作）
2. **cache_manager.py**：`match_prefix` 在 prefix caching 关闭时提前 return 前写入空 `MatchResult()`，避免调用方解引用 `_match_result=None` 崩溃
3. **transfer_manager.py**：`_build_device_layer_indices` 在 `_cache_kvs_map` 为空时也重置四个层索引列表，防止残留旧 tensor 被 swap 算子使用
4. **request.py**：`BatchRequest.append_swap_metadata` / `append_evict_metadata` 构造 `CacheSwapMetadata` 时将 `src_type`/`dst_type` 从字符串改为 `CacheLevel` 枚举，与字段类型声明一致；补充 `CacheLevel` 导入；`match_result` 属性返回类型标注修正为 `Optional[MatchResult]`
5. **resource_manager_v1.py**：`_allocate_gpu_blocks` 日志从 `INFO` 降级为 `DEBUG`，消除高频调度路径的日志噪音
6. **tests/engine/test_request.py**：同步更新 `src_type`/`dst_type` 断言为 `CacheLevel` 枚举值，补充 `CacheLevel` 导入

## Usage or Command

单元测试：
```bash
source .venv/py310/bin/activate
cd baidu/FastDeploy
python -m pytest tests/cache_manager/v1/test_cache_manager.py -v
python -m pytest tests/cache_manager/v1/test_transfer_manager.py -v
python -m pytest tests/engine/test_request.py -v
```

* [BugFix][KVCache] Fix BlockPool.allocate returns all blocks when num_blocks=0

## Motivation

当 `allocate(num_blocks=0)` 被调用时，Python 负索引陷阱导致严重错误：
`-0 == 0`，所以 `self._free_blocks[-0:]` 等价于 `self._free_blocks[0:]`，
会返回并清空整个空闲块列表，而非返回空列表。

## Modifications

在 `BlockPool.allocate` 中增加对 `num_blocks == 0` 的提前判断，直接返回 `[]`，
避免触发 Python 负索引陷阱。

## Usage or Command

```bash
# 运行相关单元测试验证修复
python -m pytest tests/cache_manager/v1/test_cache_manager.py -vv -s
```

* [KVCache][Test] add unit tests for cache_manager v1 modules

## Motivation

补全 cache_manager/v1 各模块的单测覆盖，确保核心方法有完整的测试保障。

## Modifications

新增/补充以下测试文件，全部 326 个用例通过：

- tests/cache_manager/v1/test_block_pool.py（新建）
  覆盖 BlockPool.get_metadata/set_metadata/resize、DeviceBlockPool/HostBlockPool
- tests/cache_manager/v1/test_metadata.py（新建）
  覆盖 BlockNode、RadixTreeStats、MatchResult、CacheSwapMetadata、AsyncTaskHandler
- tests/cache_manager/v1/test_cache_utils.py（补充）
  新增 hash_block_tokens、get_request_block_hasher、LayerDoneCounter 时间追踪及内部辅助方法
- tests/cache_manager/v1/test_radix_tree.py（补充）
  新增 TestCompleteSwapToDevice 专项测试类（6 个用例）
- tests/cache_manager/v1/test_cache_manager.py（补充）
  新增 offload_to_host、load_from_host、pending backup 系列、prepare_prefetch_metadata
- tests/cache_manager/v1/test_transfer_manager.py（补充）
  新增 _swap_single_layer 校验路径、sync_input/output_stream、record_input_stream_event

## Usage or Command

```bash
# 运行所有新增单测
source .venv/py310/bin/activate
python -m pytest tests/cache_manager/v1/test_block_pool.py \
  tests/cache_manager/v1/test_metadata.py \
  tests/cache_manager/v1/test_cache_utils.py \
  tests/cache_manager/v1/test_radix_tree.py \
  tests/cache_manager/v1/test_cache_manager.py \
  tests/cache_manager/v1/test_transfer_manager.py -v
# 期望结果：326 passed
```

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>
 2026-04-21 14:39:00 +08:00
freeliuzc 43685a98a7 [BugFix] Fix real token exceeding max_batched_tokens limit (#7438) 
* fix max_num_batched_tokens error compute

* add temperatory solution

* fix bug
 2026-04-17 16:18:07 +08:00
RichardWooSJTU d2d633b05c allow parallel dp starting (#7426) 2026-04-16 18:43:09 +08:00
zhouchong 6e16438a57 [Feature] implement log channel separation and request log level system (#7190) 
* feat: implement log channel separation and request log level system

* fix: log system improvements based on review

* add request_id to error logs, use RequestLogLevel enum, and unify logger implementation from utils to logger module
 2026-04-16 15:13:05 +08:00
Bingoo 6b891da02b [Optimization] enable trtllm_all_reduce fusion kernel in glm model (#6660) 
* enable trtllm_all_reduce fusion kernel in glm model

* fix conflict

* format update

* fix a bug

* modify test

* modify test

* support empty tensor and modify test

* fix test_linear config issues

* modify test name

* add edge test case

* modify format

* fix conflict

* modify default max token num in trtllm_allreduce_fusion

* add max token num branch for trtllm_allreduce_fusion

* fix format

* fix rmsnorm config issue

* modify 2025 to 2026

* using compat grard

* Lazily import flashinfer.comm and fix test config issue

* fix test issues

* add flashinfer cache dir clean machine

* fix some issues
 2026-04-16 14:10:19 +08:00
jc e53f5184ac PD deployment support without router (#7412) 2026-04-15 20:13:07 +08:00
AIbin 8eebbcaf15 [BugFix][Scheduler]Fix FD_DISABLE_CHUNKED_PREFILL max_num_batched_tokens limit (#7407) 
* fix FD_DISABLE_CHUNKED_PREFILL max_num_batched_tokens=max_model_len

* fix FD_DISABLE_CHUNKED_PREFILL max_num_batched_tokens=max_model_len
 2026-04-15 15:55:11 +08:00
Echo-Nie 8819a039c9 [Others] Fix typo (#7280) 
* typo

* typo

* typo

* typo
 2026-04-14 17:28:22 +08:00
kevin ff47701f31 [BugFix][PD Disaggregation][KVCache] Fix low cache hit rate in PD split scenario (#7364) 
## Motivation

在 PD 分离场景下，decode 节点在接收 prefill 节点转发的请求后，没有及时更新 cache block 的命中信息，
导致 prefix cache 命中率低，影响推理性能。

## Modifications

1. 在 `_free_blocks_when_stop` 方法中，额外排除 prefill 节点（`splitwise_role == "prefill"`）
   的 cache block 更新，避免 prefill 节点重复更新 cache 导致状态混乱。
2. 在 decode 节点分配请求（`_alloc_requests_with_cache`）成功后，主动调用
   `update_cache_blocks` 使用 `need_prefill_tokens` 更新 cache block 信息，
   确保 decode 节点能正确感知已命中的 prefix cache。
 2026-04-14 16:15:43 +08:00
Jiang-Jia-Jun 26d6a20c2f [Optim] Remove IPCLock between CacheManager and WorkerProcess (#7299) 
* [Optim] Remove IPCLock between CacheManager and WorkerProcess

* Update envs.py

* Update worker_process.py

---------

Co-authored-by: jiang-jia-jun <jiangjiajun@baidu.com>
 2026-04-12 13:59:34 +08:00
bukejiyu 14d46181b8 [Loader] add multi-thread model loading (#6877) 
* multi-thread-loader

* fix ut
 2026-04-09 23:40:15 -07:00
GoldPancake c1fb3112f8 [FDConfig] Support CLI args for quantization params and add cudagraph validation (#7281) 
* refactor quant cli param
 2026-04-10 14:13:42 +08:00
guozhuangzhuang 757bafe3bd [Engine][DataProcessor] fix decode token (#7102) 2026-04-08 15:41:32 +08:00
K11OntheBoat bb48bcbaa2 Split enable_mm (#7183) 
Co-authored-by: liuruian <liuruian@MacBook-Pro.local>
 2026-04-08 11:25:41 +08:00
sunxin ae2f9f4d22 [BugFix] Enable moe_gate_fp32 using FD_ENABLE_RL (#7130) 
* rl gate fp32

* clean
 2026-04-06 21:07:38 -07:00
Yonghua Li 3b8dac3b97 [BugFix] prevent requests from entering running state without a slot (#7141) 
* [fix] prevent requests from entering running state without a slot

* [fix] count abort set

* [fix] count preempted task in waiting list
 2026-04-03 14:07:57 +08:00
chenjian 2632e6cf32 [Feature] Support chunk prefill disabled in scheduler v1 (#7152) 2026-04-03 10:18:14 +08:00
Yonghua Li 98f3fc9267 [RL] [KVCache] let cache transfer managers update key prefix after weight update and add unit tests (#7083) 
* [test] add a few unit tests

* [feat] update key prefix when model weights are updated

* [test] try to fix test_worker_process
 2026-04-02 19:58:41 +08:00
sunxin c29e86fc9d [Feature] Support mtp overlap schedule (#7001) 2026-04-01 14:24:26 +08:00
zhouchong 91c832f607 [Feature] Add logging parameters and error output to terminal (#7098) 2026-04-01 13:18:42 +08:00
jc af51fc46d6 [PD Disaggregation] Write the cache of preempted req to storage and refine PD Disaggregation (#7107) 
* Write the cache of preempted req to storage

* up

* fix
 2026-04-01 13:15:52 +08:00
luukunn 3651113ee5 [DataProcessor]Remove ENABLE_V1_DATA_PROCESSOR (#7052) 
* remove ENABLE_V1_DATA_PROCESSOR

* fix unit test

* fix unit test
 2026-04-01 09:53:41 +08:00
qwes5s5 daa95244f7 abort requests (#6992) 2026-03-31 11:02:26 +08:00
chenjian 6727df8286 [Optimization] Optimize ttft for prefill pd (#6680) 
* optimize ttft

* fix

* fix

* fix ci

* fix ci

* fix

* fix bug

* fix

* add comments

* fix ci

* fix

* fix ci

* fix format

* update according to review

* add comment

* fix

* fix format
 2026-03-30 20:36:23 +08:00
freeliuzc 4fd877ed43 [Speculative Decoding] Support mtp expert-parallel and support different modality deploy (#7018) 
* support mtp ep and support different modality

* fix default arg
 2026-03-26 13:52:16 +08:00
Yonghua Li a7f52c300d [Feature] support v1 update/clear api for RL (#6761) 
* [Feature] support v1 update/clear api for RL

* [fix] fix execute_model and add sleep/wakeup api

* [fix] fix mtp and key_prefix

* [chore] move _update_key_prefix to resume method

* [fix] make the interface safe to call multiple times

* [fix] fix some tiny bugs

* [chore] make small changes against pr review

* [docs] add docs for weight update

* [test] add some tests and update docs

* [style] fix code style check

* [test] fix ci

* [fix] fix stale control responses when control method timed out

* [chore] remove unused code

* [chore] fix code style

* [chore] optimize tags and key_prefix

* [test] fix ci

* [chore] fix code style

* [test] fix ci

* [fix] fix ep control

* [fix] fix ep control for engine cache queue
 2026-03-25 19:18:46 +08:00
jc bb881c2c0a [PD Disaggregation] pd + cache_storage support vl model (#6906) 
* pd + cache_storage support vl model

* support vl model

* fix test
 2026-03-23 15:35:20 +08:00
luukunn f4a79d4c00 [Optimization]Unified data processing for online and offline (#6891) 
* remove process_request

* fix chat

* fix unit test

* remove process response

* fix unit test

* fix offline decode

* Potential fix for pull request finding

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

* fix sampling_params

---------

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
 2026-03-19 21:56:09 +08:00
luukunn c3d8db85c4 [Optimization] Update ZMQ server (#6735) 
* add batch zmq send reaponse

* update

* Revert "update"

This reverts commit 0234a25b47.

* update

* remove lock

* fix unit test

* add unit test

* add unit test

* pre commit

* add unit test

* fix unit test

* add unit test

* fix worker>1

* update zmq_worker_pid

* fix unit test

* fix unit test

* fix unit test

* add unit test

* fix unit test

* fix first token time

* fix logprobs

* add unit test

* op

* remore debug log

---------

Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>
 2026-03-19 21:53:16 +08:00
Jiang-Jia-Jun 12eb001d0c Remove comments on multi-mode request handling 
Removed comments about multi-mode scenarios and request pulling.
 2026-03-17 14:49:00 +08:00
jc 950366e58d [PD Disaggregation][RL] Register to router with version and support rdma eager connect for pd (#6718) 
* [Feature] Register to router with version info for PD disaggregation

Add RegisterManager for PD (Prefill-Decode) disaggregated deployment:
- All instances (Prefill/Decode) register to Router with heartbeat
- Prefill instances fetch Decode instance list from Router
- Prefill instances establish eager RDMA connections to Decode instances
- Register info includes: host_ip, port, role, version, is_paused, connected_decodes

Changes:
- Add RegisterManager class for managing PD registration and RDMA connections
- Add version field to ModelConfig for model version tracking
- Add connected_decodes to register_info for tracking connected Decode instances
- Add FD_ENABLE_PD_RDMA_EAGER_CONNECT environment variable

Test fixes:
- Add None checks for load_config in FDConfig.__init__
- Add version attribute to test mock model configs

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* refine

* remove test

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
 2026-03-17 14:43:35 +08:00
qwes5s5 3b7507a4c2 test_abort (#6743) 2026-03-17 14:06:40 +08:00
gongweibao a6351dea0b [BugFix][Optimization] Replace silent failures with catchable exceptions and informative error messages (#6533) 
* init

* init

* fix format

* add

* add files

* add ut

* fix some

* add ut

* add more

* add

* fix pre-commit

* fix pre-commit

* fix cover

* skip long seq

* add

* add

* fix

* remove not need

* fix set attr

* fix comments

* fix comments

* fix failed tests

---------

Co-authored-by: gongweibao <gognweibao@baidu.com>
 2026-03-16 21:32:43 +08:00
Jiang-Jia-Jun d113397b09 Simplify available_blocks assignment logic (#6819) 2026-03-16 20:12:30 +08:00
jc 04fde3b227 [PD Disaggregation] Prefill and decode support cache storage (#6768) 
* Prefill and decode support cache storage

* up

* up

* update docs and refine mooncake store

* up
 2026-03-16 14:44:49 +08:00
RichardWooSJTU 9f0778f991 [Feature] Support EP prefill with num_worst_tokens (#6574) 
* support num worst tokens

* support num worst tokens

* fix build error

* support num worst tokens: fix errors

* support num worst tokens: fix feild

* support num worst tokens: delete requiements

* replace permute and depermute op by pure cuda

* replace permute and depermute op by pure cuda

* fix ci

* fix op

* fix nan

* fix code style

---------

Co-authored-by: YuBaoku <49938469+EmmonsCurse@users.noreply.github.com>
 2026-03-11 17:09:07 +08:00
freeliuzc cf7934a4b2 [Speculative Decoding] Unify Spec and non-spec branch (#6685) 
* optimize spec-inference architecture

* delete debug log

* optimize spec_method usage  && fix unit_test

* add claude unit-test skill

* fix some ugly bug

* enhance robustness and bounds check

* unify method & spec_method to method to avoid bug

* activate CI

* fix unit test

* Unify logprobs computation for naive and speculative decoding, fix CUDA kernel

* fix logprob bug && optimize verify kernel

* fix exist_decode() judge
 2026-03-10 23:58:44 -07:00
ddchenhao66 a502dda1fe [BugFix] fix multi-step mtp bug (#6754) 2026-03-11 10:16:04 +08:00
Jiang-Jia-Jun b05a6c4206 [BugFix][KVCache] Add inter-process lock to fix NaN error under DP+EP (#6724) 
* [BugFix] Support  to fix NaN bug in EP

* Optimze notion for all the funs

* Fix potential lock contention failure issues

* Update fastdeploy/inter_communicator/ipc_signal.py

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* Update envs.py

* Update default value for USE_KVCACHE_LOCK

Change default value of USE_KVCACHE_LOCK from 1 to 0.

* Update worker_process.py

* Fix suffix wrong

* Update test_prefix_cache_manager.py

---------

Co-authored-by: Jiang-Jia-Jun <jiangjiajun@baidu.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
 2026-03-10 21:55:32 +08:00
sunxin 812657beee fix pd overlap (#6753) 2026-03-10 20:29:54 +08:00
sunxin 28f7727a3d [Feature] Set overlap schedule as default (#6668) 
* overlap default
 2026-03-09 22:34:54 +08:00
1 3a85ecf3bc [Others] Fix typos in log messages and comments (#6707) 
Fix spelling errors in log messages, docstrings, and comments:
- 'occured' -> 'occurred' (8 instances)
- 'Recieve'/'recieved' -> 'Receive'/'received' (7 instances)
- 'happend' -> 'happened' (3 instances)
- 'expet_servic' -> 'expert_service' (2 instances)
- 'meas' -> 'means' (1 instance)

No functional changes. Only log strings, docstrings, and comments are affected.

Co-authored-by: cloudforge1 <cloudforge1@users.noreply.github.com>
 2026-03-09 10:26:25 +08:00
ddchenhao66 3c0ff20328 [BugFix] fix incorrect function parameters of start_data_parallel_service (#6674) 2026-03-09 10:15:50 +08:00
SunLei 5d9524fc3c [Models][Feature] Support new ERNIE reward model and add return_token_ids to reward API (#6638) 
* reward model

* Add support for pooling-based inference in the reward model

* bugfix

---------

Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>
 2026-03-06 18:51:00 +08:00
Yonghua Li fa1906bd6f [BugFix] Fix inaccurate cache hit rate and TTFT after request preemption (#6620) 
* [chore] add has_been_rescheduled flag for requests

* [refactor] rename reschedule to preempted for accuracy and fix cache hit metrics

* [chore] add ttft_s
 2026-03-05 16:25:02 +08:00
ddchenhao66 fa4815b93a [BugFix] fix dp sheduler bug in ep4tp1 when start by using multi_api_server (#6598) 
* [BugFix] fix dp sheduler bug in ep4tp1 when start by using multi_api_server

* [BugFix] modify request_queue and result_queue of dp scheduler
 2026-03-05 10:04:12 +08:00
yzwu 3345641f4e [Iluvatar][CI] fix the dim error of seq_lens_encoder and seq_lens_decoder (#6637) 2026-03-04 14:00:40 +08:00
qwes5s5 375b5b7b21 [Feature]Log Format Normalization and Trace Log Optimization (#6370) 
* log refactor

* log refactor 2

* log refactor 3
 2026-03-03 11:31:45 +08:00
kevin 5d42f19e0a [BugFix][Scheduler] Fix can_schedule_block_num_threshold calculation (#6541) 
* fix mtp acceptance rate decline

* [BugFix][Scheduler] Fix can_schedule_block_num_threshold calculation

Fix the calculation of can_schedule_block_num_threshold in
ResourceManagerV1. The original formula using need_prefill_tokens
could lead to incorrect threshold values. Now directly use
num_chunk_new_block for accurate block scheduling.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
 2026-02-28 16:23:18 +08:00
sunxin 53aaac69da [Optimization] Enable BF16 gate computation for GLM and Qwen (#6457) 
* gate bf16

* add gate-fp32

* fix

* update baseline

* update

* update

* fix
 2026-02-26 21:08:46 -08:00