sglang icon indicating copy to clipboard operation
sglang copied to clipboard
Published 1 month ago • verified publisher icon sgl-project

[Bug]DeepSeek-R1 IndexError: index 2383 is out of bounds for dimension 0 with size 2383

Open fengyang95 opened this issue 10 months ago • 2 comments

Checklist

  • [x] 1. I have searched related issues but cannot get the expected help.
  • [x] 2. The bug has not been fixed in the latest version.
  • [x] 3. Please note that if the bug-related issue you submitted lacks corresponding environment info and a minimal reproducible demo, it will be challenging for us to reproduce and resolve the issue, reducing the likelihood of receiving feedback.
  • [x] 4. If the issue you raised is not a bug but a question, please raise a discussion at https://github.com/sgl-project/sglang/discussions/new/choose Otherwise, it will be closed.
  • [x] 5. Please use English, otherwise it will be closed.

Describe the bug

2025-02-16 19:29:38 TP4] Scheduler hit an exception: Traceback (most recent call last): File "/opt/tiger/sglang/python/sglang/srt/managers/scheduler.py", line 1827, in run_scheduler_process scheduler.event_loop_normal() File "/home/tiger/.pyenv/versions/3.11.2/lib/python3.11/site-packages/torch/utils/_contextlib.py", line 116, in decorate_context return func(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^ File "/opt/tiger/sglang/python/sglang/srt/managers/scheduler.py", line 479, in event_loop_normal self.process_batch_result(batch, result) File "/opt/tiger/sglang/python/sglang/srt/managers/scheduler.py", line 1120, in process_batch_result self.process_batch_result_prefill(batch, result) File "/opt/tiger/sglang/python/sglang/srt/managers/scheduler.py", line 1181, in process_batch_result_prefill self.tree_cache.cache_unfinished_req(req) File "/opt/tiger/sglang/python/sglang/srt/mem_cache/chunk_cache.py", line 62, in cache_unfinished_req kv_indices = self.req_to_token_pool.req_to_token[ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ IndexError: index 2383 is out of bounds for dimension 0 with size 2383

[2025-02-16 19:29:38 TP7] Scheduler hit an exception: Traceback (most recent call last): File "/opt/tiger/sglang/python/sglang/srt/managers/scheduler.py", line 1827, in run_scheduler_process scheduler.event_loop_normal() File "/home/tiger/.pyenv/versions/3.11.2/lib/python3.11/site-packages/torch/utils/_contextlib.py", line 116, in decorate_context return func(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^ File "/opt/tiger/sglang/python/sglang/srt/managers/scheduler.py", line 479, in event_loop_normal self.process_batch_result(batch, result) File "/opt/tiger/sglang/python/sglang/srt/managers/scheduler.py", line 1120, in process_batch_result self.process_batch_result_prefill(batch, result) File "/opt/tiger/sglang/python/sglang/srt/managers/scheduler.py", line 1181, in process_batch_result_prefill self.tree_cache.cache_unfinished_req(req) File "/opt/tiger/sglang/python/sglang/srt/mem_cache/chunk_cache.py", line 62, in cache_unfinished_req kv_indices = self.req_to_token_pool.req_to_token[ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ IndexError: index 2383 is out of bounds for dimension 0 with size 2383

Reproduction

two H20 Nodes

python -m sglang.launch_server --model-path /opt/tiger/DeepSeek-R1 --tp 16 --dist-init-addr ip:3894 --nnodes 2 --node-rank 0 --kv-cache-dtype fp8_e5m2 --trust-remote-code --port $PORT0 --speculative-algo NEXTN --speculative-draft /opt/tiger/DeepSeek-R1-NextN --speculative-num-steps 2 --speculative-eagle-topk 4 --speculative-num-draft-tokens 4 --disable-radix --mem-fraction-static 0.7

python -m sglang.launch_server --model-path /opt/tiger/DeepSeek-R1 --tp 16 --dist-init-addr ip:3894 --nnodes 2 --node-rank 1 --kv-cache-dtype fp8_e5m2 --trust-remote-code --port $PORT0 --speculative-algo NEXTN --speculative-draft /opt/tiger/DeepSeek-R1-NextN --speculative-num-steps 2 --speculative-eagle-topk 4 --speculative-num-draft-tokens 4 --disable-radix --mem-fraction-static 0.7

Environment

Python: 3.11.2 (main, Jul 23 2024, 17:09:09) [GCC 10.2.1 20210110] CUDA available: True GPU 0,1,2,3,4,5,6,7: NVIDIA H20 GPU 0,1,2,3,4,5,6,7 Compute Capability: 9.0 CUDA_HOME: /usr/local/cuda NVCC: Cuda compilation tools, release 12.4, V12.4.131 CUDA Driver Version: 535.183.06 PyTorch: 2.5.1+cu124 sglang: 0.4.3 sgl_kernel: 0.0.3.post6 flashinfer: 0.2.1.post1 triton: 3.1.0 transformers: 4.48.3 torchao: 0.7.0 numpy: 1.26.4 aiohttp: 3.11.12 fastapi: 0.115.8 hf_transfer: 0.1.9 huggingface_hub: 0.28.1 interegular: 0.3.3 modelscope: 1.22.0 orjson: 3.10.14 packaging: 24.2 psutil: 7.0.0 pydantic: 2.10.6 multipart: 0.0.20 zmq: 26.2.1 uvicorn: 0.34.0 uvloop: 0.21.0 vllm: 0.7.2 openai: 1.63.0 tiktoken: 0.9.0 anthropic: 0.42.0 decord: 0.6.0 NVIDIA Topology: GPU0 GPU1 GPU2 GPU3 GPU4 GPU5 GPU6 GPU7 NIC0 NIC1 NIC2 NIC3 NIC4 NIC5 NIC6 NIC7 NIC8 CPU Affinity NUMA Affinity GPU NUMA ID GPU0 X NV18 NV18 NV18 NV18 NV18 NV18 NV18 SYS PIX NODE NODE NODE SYS SYS SYS SYS 1,4-89 0N/A GPU1 NV18 X NV18 NV18 NV18 NV18 NV18 NV18 SYS NODE PIX NODE NODE SYS SYS SYS SYS 1,4-89 0N/A GPU2 NV18 NV18 X NV18 NV18 NV18 NV18 NV18 SYS NODE NODE PIX NODE SYS SYS SYS SYS 1,4-89 0N/A GPU3 NV18 NV18 NV18 X NV18 NV18 NV18 NV18 SYS NODE NODE NODE PIX SYS SYS SYS SYS 1,4-89 0N/A GPU4 NV18 NV18 NV18 NV18 X NV18 NV18 NV18 SYS SYS SYS SYS SYS PIX NODE NODE NODE 91,94-179 1 N/A GPU5 NV18 NV18 NV18 NV18 NV18 X NV18 NV18 SYS SYS SYS SYS SYS NODE PIX NODE NODE 91,94-179 1 N/A GPU6 NV18 NV18 NV18 NV18 NV18 NV18 X NV18 SYS SYS SYS SYS SYS NODE NODE PIX NODE 91,94-179 1 N/A GPU7 NV18 NV18 NV18 NV18 NV18 NV18 NV18 X SYS SYS SYS SYS SYS NODE NODE NODE PIX 91,94-179 1 N/A NIC0 SYS SYS SYS SYS SYS SYS SYS SYS X SYS SYS SYS SYS SYS SYS SYS SYS NIC1 PIX NODE NODE NODE SYS SYS SYS SYS SYS X NODE NODE NODE SYS SYS SYS SYS NIC2 NODE PIX NODE NODE SYS SYS SYS SYS SYS NODE X NODE NODE SYS SYS SYS SYS NIC3 NODE NODE PIX NODE SYS SYS SYS SYS SYS NODE NODE X NODE SYS SYS SYS SYS NIC4 NODE NODE NODE PIX SYS SYS SYS SYS SYS NODE NODE NODE X SYS SYS SYS SYS NIC5 SYS SYS SYS SYS PIX NODE NODE NODE SYS SYS SYS SYS SYS X NODE NODE NODE NIC6 SYS SYS SYS SYS NODE PIX NODE NODE SYS SYS SYS SYS SYS NODE X NODE NODE NIC7 SYS SYS SYS SYS NODE NODE PIX NODE SYS SYS SYS SYS SYS NODE NODE X NODE NIC8 SYS SYS SYS SYS NODE NODE NODE PIX SYS SYS SYS SYS SYS NODE NODE NODE X

Legend:

X = Self SYS = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI) NODE = Connection traversing PCIe as well as the interconnect between PCIe Host Bridges within a NUMA node PHB = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU) PXB = Connection traversing multiple PCIe bridges (without traversing the PCIe Host Bridge) PIX = Connection traversing at most a single PCIe bridge NV# = Connection traversing a bonded set of # NVLinks

NIC Legend:

NIC0: mlx5_0 NIC1: mlx5_1 NIC2: mlx5_2 NIC3: mlx5_3 NIC4: mlx5_4 NIC5: mlx5_5 NIC6: mlx5_6 NIC7: mlx5_7 NIC8: mlx5_8

Hypervisor vendor: KVM ulimit soft: 1024768

fengyang95 avatar Feb 16 '25 11:02 fengyang95

max_running_requests=2383

fengyang95 avatar Feb 16 '25 11:02 fengyang95

CC @ispobock

shuaills avatar Feb 16 '25 12:02 shuaills

I hit the same bug when enable NEXTN

Image

hcyz33 avatar Feb 18 '25 08:02 hcyz33

Please disable radix cache with nextn is on.

shuaills avatar Feb 18 '25 09:02 shuaills

Please disable radix cache with nextn is on. When NEXTN is enabled, it seems that the radix cache is disabled by default

Image

hcyz33 avatar Feb 20 '25 15:02 hcyz33