Docker · Workflow runs · ModelTC/LightLLM · GitHub

Actions

Docker

Actions

Loading...
Loading

docker-publish.yml

212 workflow run results

212 workflow run results

improve group_gemm kernel. (#891) Docker #524: Commit b19a832 pushed by shihaobai

8m 35s main

main

8m 35s

sgl_kernel && vllm ops (#890) Docker #523: Commit 7a18054 pushed by hiworldwzj

7m 55s main

main

7m 55s

[FIX] new whisper (#889) Docker #522: Commit f790b30 pushed by hiworldwzj

5m 53s main

main

5m 53s

vit check max batch size infer (#887) Docker #521: Commit 5d40700 pushed by hiworldwzj

5m 19s main

main

5m 19s

fix circle ref (#884) Docker #520: Commit 2e559dc pushed by shihaobai

5m 22s main

main

5m 22s

model factory (#882) Docker #519: Commit f25d1fe pushed by hiworldwzj

5m 19s main

main

5m 19s

fix: remove use_dynamic_prompt_cache code in flashinfer_struct.py (#883) Docker #518: Commit 0739a5a pushed by hiworldwzj

2m 36s main

main

2m 36s

add check for rope and tuning qwen3 on H200 (#880) Docker #517: Commit 42e8199 pushed by shihaobai

3m 23s main

main

3m 23s

fp8 scale repeat for qwen3 (#879) Docker #516: Commit b60ac4f pushed by shihaobai

2m 40s main

main

2m 40s

fix qwen3 infer bug. (#878) Docker #515: Commit c46810b pushed by shihaobai

2m 39s main

main

2m 39s

Qwen3MOE for tp8 (repeat kv) and qwen3 dense fix (#877) Docker #514: Commit f3d8e61 pushed by hiworldwzj

2m 39s main

main

2m 39s

[Feature] Add OpenAI Compatible Function Call Interface (#853) Docker #513: Commit 37e5071 pushed by hiworldwzj

4m 33s main

main

4m 33s

add qwen3 and qwen3_moe (#875) Docker #512: Commit 417b3d1 pushed by hiworldwzj

4m 48s main

main

4m 48s

[FIX] support rms_norm input is non-contiguous (#874) Docker #511: Commit 66a5e8a pushed by hiworldwzj

8m 43s main

main

8m 43s

remove b_start_loc infer input param. (#873) Docker #510: Commit aba94b6 pushed by hiworldwzj

12m 17s main

main

12m 17s

update test (#871) Docker #509: Commit 0b95787 pushed by hiworldwzj

4m 46s main

main

4m 46s

recode inferstate_struct. (#869) Docker #508: Commit e023d03 pushed by hiworldwzj

4m 21s main

main

4m 21s

fix pd split bug for multi node. (#872) Docker #507: Commit 918f01e pushed by hiworldwzj

6m 30s main

main

6m 30s

[fix]input_ids=0 (#868) Docker #506: Commit 67919d8 pushed by hiworldwzj

5m 31s main

main

5m 31s

feat: add flashinfer for llama (#859) Docker #505: Commit 3c955e6 pushed by hiworldwzj

3m 57s main

main

3m 57s

fix pd decode req classed error. (#866) Docker #504: Commit 796f11e pushed by hiworldwzj

16m 23s main

main

16m 23s

fix prefix cache of ds with fa3 (#861) Docker #503: Commit 9bbb40b pushed by hiworldwzj

18m 33s main

main

18m 33s

[Fix] adjust pip install order to keep nccl version as 2.25.1 (#863) Docker #502: Commit 3a96e3c pushed by XHPlus

20m 51s main

main

20m 51s

[Fix] use Flash-Attn-3 wheel in Dockerfile (#862) Docker #501: Commit 4f3c999 pushed by XHPlus

12m 39s main

main

12m 39s

Code clean up (#860) Docker #500: Commit 7729439 pushed by hiworldwzj

6h 0m 15s main

main

6h 0m 15s