pansicheng f84db115b1 Add storage read/write bandwidth logs to monitor kvcache performance (#9965) ...

Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu>

2025-09-05 16:52:55 -07:00

..

configs

[Fix] Add speculative_draft_model_revision to server_args (#5255)

2025-09-05 19:45:46 +08:00

connector

fix: fix MLA for ShardedModelLoader/RemoteModelLoader (#6287)

2025-08-28 16:10:09 -07:00

constrained

Update grok.py and tiktoken tokenizer (#9532)

2025-08-23 05:40:18 -07:00

debug_utils

Fix and enhance dumper (#8725)

2025-09-05 19:51:09 +08:00

disaggregation

[Vulnerability]feat(conn): set bootstrap server host (#9931)

2025-09-05 17:36:17 +08:00

distributed

Save memory for expert model parallel (#9957)

2025-09-04 13:31:47 -07:00

entrypoints

Update flashinfer to 0.3.1 for B300 support (#10087)

2025-09-05 13:41:01 -07:00

eplb

feature(eplb): add min-rebalancing-utilization-threshold for eplb (#8345)

2025-08-30 11:24:29 +08:00

function_call

Move parsers under a single folder (#9912)

2025-09-02 18:25:04 -07:00

layers

Update wave-lang to 3.7.0 and unify Wave kernel buffer options (#10069)

2025-09-05 16:01:52 -07:00

lora

Add target module validation for init adapters (#9429)

2025-08-24 20:24:50 -07:00

managers

Add storage read/write bandwidth logs to monitor kvcache performance (#9965)

2025-09-05 16:52:55 -07:00

mem_cache

Add storage read/write bandwidth logs to monitor kvcache performance (#9965)

2025-09-05 16:52:55 -07:00

metrics

Add storage read/write bandwidth logs to monitor kvcache performance (#9965)

2025-09-05 16:52:55 -07:00

model_executor

Remove mrope position sync (#9460)

2025-09-03 11:40:53 -07:00

model_loader

fix: fix MLA for ShardedModelLoader/RemoteModelLoader (#6287)

2025-08-28 16:10:09 -07:00

models

[Bug Fix] Fix Glm4vVisionBlock norm (#9884)

2025-09-05 14:20:36 -07:00

multimodal

[feat]Ascend NPU Gemma-3-12b and Gemma-3-27b support (#8909)

2025-08-31 00:25:07 -07:00

parser

Move parsers under a single folder (#9912)

2025-09-02 18:25:04 -07:00

sampling

[feat] Reduce GPU memory overhead by using weakref (#9673)

2025-08-28 01:09:06 -07:00

speculative

[feat] Add P/D attention select for draft model (#9755)

2025-09-02 22:47:23 -07:00

tokenizer

Add reasoning_effort param in TiktokenTokenizer.apply_chat_template (#9630)

2025-08-26 00:44:20 -07:00

weight_sync

Support Flatten Tensor Update Weights to speed up MOE Update Weights by 20% (#8079)

2025-08-10 16:08:59 -07:00

_custom_ops.py

[Feature] Integrate quick allreduce and select the best allreduce implementation (#6619)

2025-07-24 20:48:42 -07:00

aio_rwlock.py

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

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

bench_utils.py

[1/2] Optimizations and refactors about quant kernel (#9534)

2025-09-05 18:45:08 +08:00

constants.py

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

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

custom_op.py

[1/N] MoE Refactor: refactor select_experts (#7966)

2025-07-19 00:51:15 -07:00

hf_transformers_utils.py

[Model] Support Meituan LongCat-Flash && LongCat-Flash-MTP (#9824)

2025-08-30 23:29:21 -07:00

host_shared_memory.py

Overlapped weight offload (#8034)

2025-08-23 02:06:46 -07:00

offloader.py

Offload tensors by sharding on GPU (#9536)

2025-08-25 00:02:49 -07:00

operations_strategy.py

[5/N] MoE Refactor: Update MoE parallelism arguments (#8658)

2025-08-01 01:20:03 -07:00

operations.py

Add fp4 quantize before all-gather for Flashinfer cutlass MoE DP (max throughput) (#7667)

2025-08-15 22:08:11 -07:00

patch_torch.py

Fix torch.compile cacheing (#5259)

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

poll_based_barrier.py

Support colocating requests (#7973)

2025-07-28 22:51:49 -07:00

server_args.py

[Fix] Add speculative_draft_model_revision to server_args (#5255)

2025-09-05 19:45:46 +08:00

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

[6/N] MoE Refactor: Cleanup MoE-related configs (#8849)

2025-08-14 21:14:53 -07:00

utils.py

Tiny support setting numa nodes for different ranks (#10006)

2025-09-05 19:01:27 +08:00

warmup.py

Improve streaming, log_level, memory report, weight loading, and benchmark script (#7632)

2025-06-29 23:16:19 -07:00