[SYCL] remove global variables #7710

airMeng · 2024-06-03T08:07:10Z

Following #7566

Remaining

leave the async copy to common backend
separate GEMM files outside of ggml-sycl.c for maintainability.

AidanBeltonS · 2024-06-03T11:05:23Z

Thanks for your refactoring! I'm aware this is a draft PR
However, just an FYI I have tested this branch and see some failures that are not on the tip. Im reporting them just in case it is helpful in your work

test-backend-ops on ARC A770 GPU

  DUP(type=f32,ne=[10,10,10,1]): GGML_ASSERT: /builds/perseus-performance-libraries/llama_ci/llama.cpp/ggml-sycl.cpp:12066: src0->backend == GGML_BACKEND_TYPE_GPU
/usr/bin/bash: line 193:  9395 Aborted                 (core dumped) ./bin/test-backend-ops -b SYCL0

llama-bench

$ ./bin/llama-bench -m /models/llama-2-13b-chat.Q4_0.gguf
ggml_sycl_init: GGML_SYCL_FORCE_MMQ:   no
ggml_sycl_init: SYCL_USE_XMX: yes
ggml_sycl_init: found 3 SYCL devices:
get_memory_info: [warning] ext_intel_free_memory is not supported (export/set ZES_ENABLE_SYSMAN=1 to support), use total memory as free memory
get_memory_info: [warning] ext_intel_free_memory is not supported (export/set ZES_ENABLE_SYSMAN=1 to support), use total memory as free memory
WARNING: /proc/sys/kernel/numa_balancing is enabled, this has been observed to impair performance
| model                          |       size |     params | backend    | ngl |          test |              t/s |
| ------------------------------ | ---------: | ---------: | ---------- | --: | ------------: | ---------------: |
[SYCL] call ggml_check_sycl
ggml_check_sycl: GGML_SYCL_DEBUG: 0
ggml_check_sycl: GGML_SYCL_F16: yes
found 3 SYCL devices:
|  |                   |                                       |       |Max    |        |Max  |Global |                     |
|  |                   |                                       |       |compute|Max work|sub  |mem    |                     |
|ID|        Device Type|                                   Name|Version|units  |group   |group|size   |       Driver version|
|--|-------------------|---------------------------------------|-------|-------|--------|-----|-------|---------------------|
| 0|       [cuda:gpu:0]|                  NVIDIA A100-PCIE-40GB|    8.0|    108|    1024|   32| 42298M|            CUDA 12.2|
| 1|     [opencl:cpu:0]|     Intel Xeon Gold 6326 CPU @ 2.90GHz|    3.0|     32|    8192|   64| 67116M|2024.17.3.0.08_160000|
| 2|     [opencl:acc:0]|            Intel FPGA Emulation Device|    1.2|     32|67108864|   64| 67116M|2024.17.3.0.08_160000|
GGML_ASSERT: /llama.cpp/ggml-sycl.cpp:13216: tensor->backend == GGML_BACKEND_TYPE_GPU
/usr/bin/bash: line 161:   230 Aborted                 (core dumped) ./bin/llama-bench -m /models/llama-2-13b-chat.Q4_0.gguf

airMeng · 2024-06-03T11:56:51Z

@AidanBeltonS Thank you for your reminder! I am aware current interaction between SYCL and common is not perfect, you can review the rough design.

github-actions · 2024-06-03T21:03:12Z

📈 llama.cpp server for bench-server-baseline on Standard_NC4as_T4_v3 for phi-2-q4_0: 542 iterations 🚀

Expand details for performance related PR only

Concurrent users: 8, duration: 10m
HTTP request : avg=8624.57ms p(95)=21737.71ms fails=, finish reason: stop=483 truncated=59
Prompt processing (pp): avg=97.04tk/s p(95)=464.26tk/s
Token generation (tg): avg=32.97tk/s p(95)=48.81tk/s
ggml-org/models/phi-2/ggml-model-q4_0.gguf parallel=8 ctx-size=16384 ngl=33 batch-size=2048 ubatch-size=256 pp=1024 pp+tg=2048 branch=sycl-remove-global-variables commit=f32f17a781e908df90e184df08183a936139fcbf

More

---
config:
    xyChart:
        titleFontSize: 12
        width: 900
        height: 600
    themeVariables:
        xyChart:
            titleColor: "#000000"
---
xychart-beta
    title "llama.cpp bench-server-baseline on Standard_NC4as_T4_v3
 duration=10m 542 iterations"
    y-axis "llamacpp:prompt_tokens_seconds"
    x-axis "llamacpp:prompt_tokens_seconds" 1717592545 --> 1717593165
    line [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 569.4, 569.4, 569.4, 569.4, 569.4, 745.11, 745.11, 745.11, 745.11, 745.11, 743.42, 743.42, 743.42, 743.42, 743.42, 760.01, 760.01, 760.01, 760.01, 760.01, 818.26, 818.26, 818.26, 818.26, 818.26, 843.48, 843.48, 843.48, 843.48, 843.48, 839.06, 839.06, 839.06, 839.06, 839.06, 862.82, 862.82, 862.82, 862.82, 862.82, 867.58, 867.58, 867.58, 867.58, 867.58, 877.97, 877.97, 877.97, 877.97, 877.97, 895.8, 895.8, 895.8, 895.8, 895.8, 858.24, 858.24, 858.24, 858.24, 858.24, 853.14, 853.14, 853.14, 853.14, 853.14, 805.84, 805.84, 805.84, 805.84, 805.84, 809.44, 809.44, 809.44, 809.44, 809.44, 811.59, 811.59, 811.59, 811.59, 811.59, 811.25, 811.25, 811.25, 811.25, 811.25, 836.07, 836.07, 836.07, 836.07, 836.07, 834.37, 834.37, 834.37, 834.37, 834.37, 834.17, 834.17, 834.17, 834.17, 834.17, 840.86, 840.86, 840.86, 840.86, 840.86, 836.29, 836.29, 836.29, 836.29, 836.29, 839.19, 839.19, 839.19, 839.19, 839.19, 843.97, 843.97, 843.97, 843.97, 843.97, 846.7, 846.7, 846.7, 846.7, 846.7, 847.92, 847.92, 847.92, 847.92, 847.92, 854.72, 854.72, 854.72, 854.72, 854.72, 854.78, 854.78, 854.78, 854.78, 854.78, 855.38, 855.38, 855.38, 855.38, 855.38, 860.3, 860.3, 860.3, 860.3, 860.3, 857.94, 857.94, 857.94, 857.94, 857.94, 856.36, 856.36, 856.36, 856.36, 856.36, 859.46, 859.46, 859.46, 859.46, 859.46, 869.63, 869.63, 869.63, 869.63, 869.63, 875.93, 875.93, 875.93, 875.93, 875.93, 882.06, 882.06, 882.06, 882.06, 882.06, 882.11, 882.11, 882.11, 882.11, 882.11, 881.93, 881.93, 881.93, 881.93, 881.93, 883.8, 883.8, 883.8, 883.8, 883.8, 885.46, 885.46, 885.46, 885.46, 885.46, 895.84, 895.84, 895.84, 895.84, 895.84, 899.44, 899.44, 899.44, 899.44, 899.44, 898.95, 898.95, 898.95, 898.95, 898.95, 896.63, 896.63, 896.63, 896.63, 896.63, 894.49, 894.49, 894.49, 894.49, 894.49, 892.86, 892.86, 892.86, 892.86, 892.86, 896.26, 896.26, 896.26, 896.26, 896.26, 897.62, 897.62, 897.62, 897.62, 897.62, 899.19, 899.19, 899.19, 899.19, 899.19, 901.08, 901.08, 901.08, 901.08, 901.08, 900.1, 900.1, 900.1, 900.1, 900.1, 899.77, 899.77, 899.77, 899.77, 899.77, 895.79, 895.79, 895.79, 895.79, 895.79, 894.72, 894.72, 894.72, 894.72, 894.72, 899.98, 899.98, 899.98, 899.98, 899.98, 899.71, 899.71, 899.71, 899.71, 899.71, 899.05, 899.05, 899.05, 899.05, 899.05, 899.19, 899.19, 899.19, 899.19, 899.19, 899.35, 899.35, 899.35, 899.35, 899.35, 899.65, 899.65, 899.65, 899.65, 899.65, 899.75]

More

---
config:
    xyChart:
        titleFontSize: 12
        width: 900
        height: 600
    themeVariables:
        xyChart:
            titleColor: "#000000"
---
xychart-beta
    title "llama.cpp bench-server-baseline on Standard_NC4as_T4_v3
 duration=10m 542 iterations"
    y-axis "llamacpp:predicted_tokens_seconds"
    x-axis "llamacpp:predicted_tokens_seconds" 1717592545 --> 1717593165
    line [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 47.38, 47.38, 47.38, 47.38, 47.38, 40.13, 40.13, 40.13, 40.13, 40.13, 27.59, 27.59, 27.59, 27.59, 27.59, 26.94, 26.94, 26.94, 26.94, 26.94, 28.33, 28.33, 28.33, 28.33, 28.33, 29.62, 29.62, 29.62, 29.62, 29.62, 31.39, 31.39, 31.39, 31.39, 31.39, 32.05, 32.05, 32.05, 32.05, 32.05, 32.91, 32.91, 32.91, 32.91, 32.91, 33.19, 33.19, 33.19, 33.19, 33.19, 33.25, 33.25, 33.25, 33.25, 33.25, 33.78, 33.78, 33.78, 33.78, 33.78, 33.4, 33.4, 33.4, 33.4, 33.4, 32.77, 32.77, 32.77, 32.77, 32.77, 31.48, 31.48, 31.48, 31.48, 31.48, 31.03, 31.03, 31.03, 31.03, 31.03, 30.59, 30.59, 30.59, 30.59, 30.59, 30.56, 30.56, 30.56, 30.56, 30.56, 30.04, 30.04, 30.04, 30.04, 30.04, 29.6, 29.6, 29.6, 29.6, 29.6, 29.6, 29.6, 29.6, 29.6, 29.6, 29.71, 29.71, 29.71, 29.71, 29.71, 30.04, 30.04, 30.04, 30.04, 30.04, 30.02, 30.02, 30.02, 30.02, 30.02, 30.24, 30.24, 30.24, 30.24, 30.24, 30.53, 30.53, 30.53, 30.53, 30.53, 30.42, 30.42, 30.42, 30.42, 30.42, 30.54, 30.54, 30.54, 30.54, 30.54, 30.85, 30.85, 30.85, 30.85, 30.85, 31.05, 31.05, 31.05, 31.05, 31.05, 31.16, 31.16, 31.16, 31.16, 31.16, 31.27, 31.27, 31.27, 31.27, 31.27, 31.5, 31.5, 31.5, 31.5, 31.5, 31.53, 31.53, 31.53, 31.53, 31.53, 31.22, 31.22, 31.22, 31.22, 31.22, 31.17, 31.17, 31.17, 31.17, 31.17, 30.86, 30.86, 30.86, 30.86, 30.86, 31.0, 31.0, 31.0, 31.0, 31.0, 31.23, 31.23, 31.23, 31.23, 31.23, 31.26, 31.26, 31.26, 31.26, 31.26, 31.45, 31.45, 31.45, 31.45, 31.45, 31.28, 31.28, 31.28, 31.28, 31.28, 31.09, 31.09, 31.09, 31.09, 31.09, 30.74, 30.74, 30.74, 30.74, 30.74, 29.36, 29.36, 29.36, 29.36, 29.36, 29.1, 29.1, 29.1, 29.1, 29.1, 29.13, 29.13, 29.13, 29.13, 29.13, 29.14, 29.14, 29.14, 29.14, 29.14, 29.3, 29.3, 29.3, 29.3, 29.3, 29.29, 29.29, 29.29, 29.29, 29.29, 29.32, 29.32, 29.32, 29.32, 29.32, 29.33, 29.33, 29.33, 29.33, 29.33, 29.24, 29.24, 29.24, 29.24, 29.24, 29.2, 29.2, 29.2, 29.2, 29.2, 29.17, 29.17, 29.17, 29.17, 29.17, 29.13, 29.13, 29.13, 29.13, 29.13, 29.24, 29.24, 29.24, 29.24, 29.24, 29.35, 29.35, 29.35, 29.35, 29.35, 29.42, 29.42, 29.42, 29.42, 29.42, 29.54, 29.54, 29.54, 29.54, 29.54, 29.64]

Details

More

---
config:
    xyChart:
        titleFontSize: 12
        width: 900
        height: 600
    themeVariables:
        xyChart:
            titleColor: "#000000"
---
xychart-beta
    title "llama.cpp bench-server-baseline on Standard_NC4as_T4_v3
 duration=10m 542 iterations"
    y-axis "llamacpp:kv_cache_usage_ratio"
    x-axis "llamacpp:kv_cache_usage_ratio" 1717592545 --> 1717593165
    line [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.2, 0.2, 0.2, 0.2, 0.2, 0.37, 0.37, 0.37, 0.37, 0.37, 0.3, 0.3, 0.3, 0.3, 0.3, 0.09, 0.09, 0.09, 0.09, 0.09, 0.19, 0.19, 0.19, 0.19, 0.19, 0.14, 0.14, 0.14, 0.14, 0.14, 0.08, 0.08, 0.08, 0.08, 0.08, 0.14, 0.14, 0.14, 0.14, 0.14, 0.18, 0.18, 0.18, 0.18, 0.18, 0.2, 0.2, 0.2, 0.2, 0.2, 0.11, 0.11, 0.11, 0.11, 0.11, 0.3, 0.3, 0.3, 0.3, 0.3, 0.27, 0.27, 0.27, 0.27, 0.27, 0.41, 0.41, 0.41, 0.41, 0.41, 0.31, 0.31, 0.31, 0.31, 0.31, 0.28, 0.28, 0.28, 0.28, 0.28, 0.18, 0.18, 0.18, 0.18, 0.18, 0.31, 0.31, 0.31, 0.31, 0.31, 0.32, 0.32, 0.32, 0.32, 0.32, 0.13, 0.13, 0.13, 0.13, 0.13, 0.22, 0.22, 0.22, 0.22, 0.22, 0.16, 0.16, 0.16, 0.16, 0.16, 0.12, 0.12, 0.12, 0.12, 0.12, 0.09, 0.09, 0.09, 0.09, 0.09, 0.12, 0.12, 0.12, 0.12, 0.12, 0.17, 0.17, 0.17, 0.17, 0.17, 0.11, 0.11, 0.11, 0.11, 0.11, 0.11, 0.11, 0.11, 0.11, 0.11, 0.12, 0.12, 0.12, 0.12, 0.12, 0.14, 0.14, 0.14, 0.14, 0.14, 0.11, 0.11, 0.11, 0.11, 0.11, 0.16, 0.16, 0.16, 0.16, 0.16, 0.15, 0.15, 0.15, 0.15, 0.15, 0.32, 0.32, 0.32, 0.32, 0.32, 0.23, 0.23, 0.23, 0.23, 0.23, 0.28, 0.28, 0.28, 0.28, 0.28, 0.07, 0.07, 0.07, 0.07, 0.07, 0.12, 0.12, 0.12, 0.12, 0.12, 0.16, 0.16, 0.16, 0.16, 0.16, 0.14, 0.14, 0.14, 0.14, 0.14, 0.34, 0.34, 0.34, 0.34, 0.34, 0.55, 0.55, 0.55, 0.55, 0.55, 0.61, 0.61, 0.61, 0.61, 0.61, 0.62, 0.62, 0.62, 0.62, 0.62, 0.34, 0.34, 0.34, 0.34, 0.34, 0.13, 0.13, 0.13, 0.13, 0.13, 0.2, 0.2, 0.2, 0.2, 0.2, 0.14, 0.14, 0.14, 0.14, 0.14, 0.16, 0.16, 0.16, 0.16, 0.16, 0.17, 0.17, 0.17, 0.17, 0.17, 0.24, 0.24, 0.24, 0.24, 0.24, 0.2, 0.2, 0.2, 0.2, 0.2, 0.29, 0.29, 0.29, 0.29, 0.29, 0.18, 0.18, 0.18, 0.18, 0.18, 0.27, 0.27, 0.27, 0.27, 0.27, 0.11, 0.11, 0.11, 0.11, 0.11, 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.12, 0.15, 0.15, 0.15, 0.15, 0.15, 0.1, 0.1, 0.1, 0.1, 0.1, 0.14]

More

---
config:
    xyChart:
        titleFontSize: 12
        width: 900
        height: 600
    themeVariables:
        xyChart:
            titleColor: "#000000"
---
xychart-beta
    title "llama.cpp bench-server-baseline on Standard_NC4as_T4_v3
 duration=10m 542 iterations"
    y-axis "llamacpp:requests_processing"
    x-axis "llamacpp:requests_processing" 1717592545 --> 1717593165
    line [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 6.0, 6.0, 6.0, 6.0, 6.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 5.0, 5.0, 5.0, 5.0, 5.0, 7.0, 7.0, 7.0, 7.0, 7.0, 5.0, 5.0, 5.0, 5.0, 5.0, 6.0, 6.0, 6.0, 6.0, 6.0, 3.0, 3.0, 3.0, 3.0, 3.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 3.0, 3.0, 3.0, 3.0, 3.0, 5.0, 5.0, 5.0, 5.0, 5.0, 7.0, 7.0, 7.0, 7.0, 7.0, 7.0, 7.0, 7.0, 7.0, 7.0, 5.0, 5.0, 5.0, 5.0, 5.0, 8.0, 8.0, 8.0, 8.0, 8.0, 4.0, 4.0, 4.0, 4.0, 4.0, 8.0, 8.0, 8.0, 8.0, 8.0, 5.0, 5.0, 5.0, 5.0, 5.0, 6.0, 6.0, 6.0, 6.0, 6.0, 8.0, 8.0, 8.0, 8.0, 8.0, 3.0, 3.0, 3.0, 3.0, 3.0, 2.0, 2.0, 2.0, 2.0, 2.0, 7.0, 7.0, 7.0, 7.0, 7.0, 5.0, 5.0, 5.0, 5.0, 5.0, 4.0, 4.0, 4.0, 4.0, 4.0, 6.0, 6.0, 6.0, 6.0, 6.0, 5.0, 5.0, 5.0, 5.0, 5.0, 6.0, 6.0, 6.0, 6.0, 6.0, 3.0, 3.0, 3.0, 3.0, 3.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 5.0, 5.0, 5.0, 5.0, 5.0, 8.0, 8.0, 8.0, 8.0, 8.0, 5.0, 5.0, 5.0, 5.0, 5.0, 8.0, 8.0, 8.0, 8.0, 8.0, 4.0, 4.0, 4.0, 4.0, 4.0, 3.0, 3.0, 3.0, 3.0, 3.0, 3.0, 3.0, 3.0, 3.0, 3.0, 6.0, 6.0, 6.0, 6.0, 6.0, 7.0, 7.0, 7.0, 7.0, 7.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 8.0, 6.0, 6.0, 6.0, 6.0, 6.0, 6.0, 6.0, 6.0, 6.0, 6.0, 2.0, 2.0, 2.0, 2.0, 2.0, 4.0, 4.0, 4.0, 4.0, 4.0, 3.0, 3.0, 3.0, 3.0, 3.0, 7.0, 7.0, 7.0, 7.0, 7.0, 5.0, 5.0, 5.0, 5.0, 5.0, 7.0, 7.0, 7.0, 7.0, 7.0, 5.0, 5.0, 5.0, 5.0, 5.0, 6.0, 6.0, 6.0, 6.0, 6.0, 8.0, 8.0, 8.0, 8.0, 8.0, 4.0, 4.0, 4.0, 4.0, 4.0, 1.0, 1.0, 1.0, 1.0, 1.0, 7.0, 7.0, 7.0, 7.0, 7.0, 6.0, 6.0, 6.0, 6.0, 6.0, 5.0, 5.0, 5.0, 5.0, 5.0, 3.0]

airMeng · 2024-06-05T12:42:55Z

@AidanBeltonS @NeoZhangJianyu now main.exe can work now. However, there seems to be huge performance regression which I am working on. You can review the design part now.

NeoZhangJianyu · 2024-06-07T06:17:58Z

@AidanBeltonS @NeoZhangJianyu now main.exe can work now. However, there seems to be huge performance regression which I am working on. You can review the design part now.

Yes.
I think this PR should reorganize the code. It shouldn't impact the performance.

AidanBeltonS

Thanks for the changes! This is a big but necessary update. I have some minor comments

ggml-sycl.cpp

remove global variables and pack into context

NeoZhangJianyu · 2024-06-13T06:26:19Z

Still crash in multiple GPUs:

airMeng · 2024-06-14T09:38:01Z

@AidanBeltonS @NeoZhangJianyu This PR fixed SYCL broken since #7640 (comment) and I believe it solves #7777 and related, please have a try.

Known issues: multi-card support still broken

AidanBeltonS · 2024-06-14T15:39:25Z

@AidanBeltonS @NeoZhangJianyu This PR fixed SYCL broken since #7640 (comment) and I believe it solves #7777 and related, please have a try.

I have tested on the A100 GPU and can confirm this fixes #7777

airMeng marked this pull request as draft June 3, 2024 08:07

airMeng added refactoring Refactoring Intel GPU Review Complexity : High Generally require indepth knowledge of LLMs or GPUs labels Jun 3, 2024

github-actions bot added build Compilation issues ggml changes relating to the ggml tensor library for machine learning SYCL https://en.wikipedia.org/wiki/SYCL - GPU programming language labels Jun 3, 2024

airMeng requested a review from NeoZhangJianyu June 4, 2024 08:52

airMeng force-pushed the sycl-remove-global-variables branch from b49f1c0 to 8dfc5a7 Compare June 5, 2024 12:32

airMeng marked this pull request as ready for review June 5, 2024 12:36

airMeng force-pushed the sycl-remove-global-variables branch from cc8c48b to f32f17a Compare June 5, 2024 12:41

AidanBeltonS approved these changes Jun 7, 2024

View reviewed changes

ggml-sycl.cpp Outdated Show resolved Hide resolved

ggml-sycl.cpp Outdated Show resolved Hide resolved

airMeng commented Jun 13, 2024

View reviewed changes

ggml-sycl.cpp Show resolved Hide resolved

airMeng added 7 commits June 13, 2024 11:39

seperate DPCT helpers outside

c29eccc

remove global variables and pack into context

replace global variables with context[2/2]

d0186d3

remove useless extra

2a034d2

update mul_mat condition

abe11fe

remove duplicate buft initialization

9c5476e

remove duplicate extra and global work group size

224273e

trim white space

d342abc

airMeng force-pushed the sycl-remove-global-variables branch from f32f17a to d342abc Compare June 13, 2024 03:49

airMeng added 2 commits June 14, 2024 09:23

remove useless backend check

996b35a

remove duplicated extras

c776fb8

airMeng mentioned this pull request Jun 15, 2024

sycl: always set the main device after initialization #7909

Closed

4 tasks

use macro for group_size and remove cuda-related

28eaafc

airMeng merged commit 7b2f4a7 into master Jun 15, 2024
68 checks passed

airMeng deleted the sycl-remove-global-variables branch June 15, 2024 06:05

This was referenced Jun 16, 2024

add ggml-sycl to sync script ggerganov/ggml#857

Merged

[SYCL] remove global variables (#7710) ggerganov/ggml#858

Merged

qnixsynapse mentioned this pull request Jun 17, 2024

[SYCL] Update llama.cpp to latest commit for fixing currently broken SYCL backend abetlen/llama-cpp-python#1536

Open

airMeng mentioned this pull request Jun 18, 2024

[SYCL] refactor #6408

Merged

2 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[SYCL] remove global variables #7710

[SYCL] remove global variables #7710

airMeng commented Jun 3, 2024 •

edited

AidanBeltonS commented Jun 3, 2024

airMeng commented Jun 3, 2024

github-actions bot commented Jun 3, 2024 •

edited

airMeng commented Jun 5, 2024

NeoZhangJianyu commented Jun 7, 2024

AidanBeltonS left a comment

NeoZhangJianyu commented Jun 13, 2024

airMeng commented Jun 14, 2024

AidanBeltonS commented Jun 14, 2024

[SYCL] remove global variables #7710

[SYCL] remove global variables #7710

Conversation

airMeng commented Jun 3, 2024 • edited

AidanBeltonS commented Jun 3, 2024

airMeng commented Jun 3, 2024

github-actions bot commented Jun 3, 2024 • edited

airMeng commented Jun 5, 2024

NeoZhangJianyu commented Jun 7, 2024

AidanBeltonS left a comment

Choose a reason for hiding this comment

NeoZhangJianyu commented Jun 13, 2024

airMeng commented Jun 14, 2024

AidanBeltonS commented Jun 14, 2024

airMeng commented Jun 3, 2024 •

edited

github-actions bot commented Jun 3, 2024 •

edited