u4lr451 ed0a0b692c Perormance: Enable cuda graph for dp idle batch (#7269) ...

Co-authored-by: austindeng <austindeng@tencent.com>
Co-authored-by: Cheng Wan <54331508+ch-wan@users.noreply.github.com>
Co-authored-by: ch-wan <cwan39@gatech.edu>

2025-06-23 17:34:13 -07:00

..

configs

Feat/support rerank (#6058)

2025-06-16 10:50:01 -07:00

connector

feat(remote_model): support variable remote backend for model loader (#3964)

2025-03-14 00:40:44 -07:00

constrained

Correctly abort the failed grammar requests & Improve the handling of abort (#6803)

2025-06-01 19:00:07 -07:00

disaggregation

[minor] simplify the TokenToKVPoolAllocator (#7414)

2025-06-22 12:37:18 +08:00

distributed

Fix All-Gather under world size one (#7219)

2025-06-20 14:57:34 -07:00

entrypoints

feat(oai refactor): Replace openai_api with entrypoints/openai (#7351)

2025-06-21 13:21:06 -07:00

eplb_simulator

Minor add utility to read expert distribution recorder output (#7134)

2025-06-12 16:50:05 -07:00

function_call

feat(func_call): Add more check in BaseFormatDetector.parse_streaming_increment (#7479)

2025-06-23 11:15:47 -07:00

layers

Perormance: Enable cuda graph for dp idle batch (#7269)

2025-06-23 17:34:13 -07:00

lora

Refactor LoRAManager and LoRAMemoryPool state management logic for dynamic LoRA loading support (#7412)

2025-06-21 16:09:19 -07:00

managers

Perormance: Enable cuda graph for dp idle batch (#7269)

2025-06-23 17:34:13 -07:00

mem_cache

[minor] simplify the TokenToKVPoolAllocator (#7414)

2025-06-22 12:37:18 +08:00

metrics

Correctly abort the failed grammar requests & Improve the handling of abort (#6803)

2025-06-01 19:00:07 -07:00

model_executor

Perormance: Enable cuda graph for dp idle batch (#7269)

2025-06-23 17:34:13 -07:00

model_loader

Support weight loading without mmap (#7469)

2025-06-23 15:13:59 -07:00

models

Quick fix for DeepGemm requant to also cover MTP. (#7378)

2025-06-23 12:08:54 -07:00

sampling

[Feature] Add Logit Bias (#6579)

2025-06-10 15:39:25 -07:00

speculative

Perormance: Enable cuda graph for dp idle batch (#7269)

2025-06-23 17:34:13 -07:00

_custom_ops.py

support qwen2 running on ascend npu device (#7022)

2025-06-17 11:24:10 -07:00

aio_rwlock.py

A better aio rwlock that guarantees the order (#2547)

2024-12-22 15:44:32 -08:00

code_completion_parser.py

feat(oai refactor): Replace openai_api with entrypoints/openai (#7351)

2025-06-21 13:21:06 -07:00

constants.py

Multi-Stage Awake: Support Resume and Pause KV Cache and Weights separately (#7099)

2025-06-19 00:56:37 -07:00

conversation.py

feat(oai refactor): Replace openai_api with entrypoints/openai (#7351)

2025-06-21 13:21:06 -07:00

custom_op.py

Update python API of activation, topk, norm and rope and remove vllm dependency (#6614)

2025-06-17 22:11:50 -07:00

debug_utils.py

Add simple utility to dump tensors for debugging (#6815)

2025-06-02 11:15:31 -07:00

hf_transformers_utils.py

[New Model] Devstral support (#6547)

2025-05-26 19:27:48 -07:00

jinja_template_utils.py

feat(oai refactor): Replace openai_api with entrypoints/openai (#7351)

2025-06-21 13:21:06 -07:00

mm_utils.py

refactor: minor refactors regarding multimodal processing (#6187)

2025-05-17 22:53:20 -07:00

model_parallel.py

Make torch TP composable with torchao (#2436)

2024-12-11 04:21:42 -08:00

operations_strategy.py

Overlap two kernels in DeepSeek with communication (#6711)

2025-05-28 15:53:51 -07:00

operations.py

Support overlapping two batches (#4068)

2025-05-24 17:39:07 -07:00

patch_torch.py

Fix torch.compile cacheing (#5259)

2025-04-10 18:08:45 -07:00

reasoning_parser.py

feat(oai refactor): Replace openai_api with entrypoints/openai (#7351)

2025-06-21 13:21:06 -07:00

server_args.py

Support weight loading without mmap (#7469)

2025-06-23 15:13:59 -07:00

torch_memory_saver_adapter.py

Multi-Stage Awake: Support Resume and Pause KV Cache and Weights separately (#7099)

2025-06-19 00:56:37 -07:00

two_batch_overlap.py

Refactor DeepGEMM integration (#7150)

2025-06-13 20:41:03 -07:00

utils.py

Fix a bug in BatchTokenIDOut & Misc style and dependency updates (#7457)

2025-06-23 06:20:39 -07:00

warmup.py

Support penalty in overlap mode; return logprob with chunked prefill; improve benchmark scripts (#3988)

2025-03-03 00:12:04 -08:00