ECE 408 sem project

This commit is contained in:
Aryan Singh committed 2026-06-07 13:43:20 -05:00
1 parent 1d79ccd2e3
commit fa3efcaf07
139 files changed
+11634 -1

No files matched your search

+108
View File
@@ -0,0 +1,108 @@
test-gpt2-kernels - TF32 Mode is turned OFF.
test-gpt2 - [State]
Batch Size: 4
Sequence Length: 64
GPT2 - Allocated 221 MiB for activations
test-gpt2 - Printing 10 Expected vs Actual Logits:
-43.431618, -43.379311
-39.836346, -39.790001
-43.065910, -43.020863
-42.828045, -42.780563
-43.529541, -43.479824
-44.318398, -44.268772
-41.227425, -41.177418
-41.270760, -41.221939
-42.541393, -42.488613
-42.394997, -42.345592
---------------- Logits OK! (max err 0.055688, RMSE 0.001083) ----------------
_________________________ All tests passed! _________________________
Collecting data...
Generating '/tmp/nsys-report-c915.qdstrm'
␍[1/8] [0% ] report6.nsys-rep␍[1/8] [0% ] report6.nsys-rep␍[1/8] [==19% ] report6.nsys-rep␍[1/8] [==21% ] report6.nsys-rep␍[1/8] [===22% ] report6.nsys-rep␍[1/8] [======35% ] report6.nsys-rep␍[1/8] [=======36% ] report6.nsys-rep␍[1/8] [==========49% ] report6.nsys-rep␍[1/8] [===========50% ] report6.nsys-rep␍[1/8] [===========52% ] report6.nsys-rep␍[1/8] [===========53% ] report6.nsys-rep␍[1/8] [============54% ] report6.nsys-rep␍[1/8] [============55% ] report6.nsys-rep␍[1/8] [============56% ] report6.nsys-rep␍[1/8] [============57% ] report6.nsys-rep␍[1/8] [=============58% ] report6.nsys-rep␍[1/8] [=============59% ] report6.nsys-rep␍[1/8] [==================75% ] report6.nsys-rep␍[1/8] [======================91% ] report6.nsys-rep␍[1/8] [========================100%] report6.nsys-rep␍[1/8] [========================100%] report6.nsys-rep
[2/8] [0% ] report6.sqlite [2/8] [1% ] report6.sqlite [2/8] [2% ] report6.sqlite [2/8] [3% ] report6.sqlite [2/8] [4% ] report6.sqlite [2/8] [5% ] report6.sqlite [2/8] [6% ] report6.sqlite [2/8] [7% ] report6.sqlite [2/8] [8% ] report6.sqlite [2/8] [9% ] report6.sqlite [2/8] [10% ] report6.sqlite [2/8] [11% ] report6.sqlite [2/8] [12% ] report6.sqlite [2/8] [13% ] report6.sqlite [2/8] [14% ] report6.sqlite [2/8] [=15% ] report6.sqlite [2/8] [=16% ] report6.sqlite [2/8] [=17% ] report6.sqlite [2/8] [==18% ] report6.sqlite [2/8] [==19% ] report6.sqlite [2/8] [==20% ] report6.sqlite [2/8] [==21% ] report6.sqlite [2/8] [===22% ] report6.sqlite [2/8] [===23% ] report6.sqlite [2/8] [===24% ] report6.sqlite [2/8] [====25% ] report6.sqlite [2/8] [====26% ] report6.sqlite [2/8] [====27% ] report6.sqlite [2/8] [====28% ] report6.sqlite [2/8] [=====29% ] report6.sqlite [2/8] [=====30% ] report6.sqlite [2/8] [=====31% ] report6.sqlite [2/8] [=====32% ] report6.sqlite [2/8] [======33% ] report6.sqlite [2/8] [======34% ] report6.sqlite [2/8] [======35% ] report6.sqlite [2/8] [=======36% ] report6.sqlite [2/8] [=======37% ] report6.sqlite [2/8] [=======38% ] report6.sqlite [2/8] [=======39% ] report6.sqlite [2/8] [========40% ] report6.sqlite [2/8] [========41% ] report6.sqlite [2/8] [========42% ] report6.sqlite [2/8] [=========43% ] report6.sqlite [2/8] [=========44% ] report6.sqlite [2/8] [=========45% ] report6.sqlite [2/8] [=========46% ] report6.sqlite [2/8] [==========47% ] report6.sqlite [2/8] [==========48% ] report6.sqlite [2/8] [==========49% ] report6.sqlite [2/8] [===========50% ] report6.sqlite [2/8] [===========51% ] report6.sqlite [2/8] [===========52% ] report6.sqlite [2/8] [===========53% ] report6.sqlite [2/8] [============54% ] report6.sqlite [2/8] [============55% ] report6.sqlite [2/8] [============56% ] report6.sqlite [2/8] [============57% ] report6.sqlite [2/8] [=============58% ] report6.sqlite [2/8] [=============59% ] report6.sqlite [2/8] [=============60% ] report6.sqlite [2/8] [==============61% ] report6.sqlite [2/8] [==============62% ] report6.sqlite [2/8] [==============63% ] report6.sqlite [2/8] [==============64% ] report6.sqlite [2/8] [===============65% ] report6.sqlite [2/8] [===============66% ] report6.sqlite [2/8] [===============67% ] report6.sqlite [2/8] [================68% ] report6.sqlite [2/8] [================69% ] report6.sqlite [2/8] [================70% ] report6.sqlite [2/8] [================71% ] report6.sqlite [2/8] [=================72% ] report6.sqlite [2/8] [=================73% ] report6.sqlite [2/8] [=================74% ] report6.sqlite [2/8] [==================75% ] report6.sqlite [2/8] [==================76% ] report6.sqlite [2/8] [==================77% ] report6.sqlite [2/8] [==================78% ] report6.sqlite [2/8] [===================79% ] report6.sqlite [2/8] [===================80% ] report6.sqlite [2/8] [===================81% ] report6.sqlite [2/8] [===================82% ] report6.sqlite [2/8] [====================83% ] report6.sqlite [2/8] [====================84% ] report6.sqlite [2/8] [====================85% ] report6.sqlite [2/8] [=====================86% ] report6.sqlite [2/8] [=====================87% ] report6.sqlite [2/8] [=====================88% ] report6.sqlite [2/8] [=====================89% ] report6.sqlite [2/8] [======================90% ] report6.sqlite [2/8] [======================91% ] report6.sqlite [2/8] [======================92% ] report6.sqlite [2/8] [=======================93% ] report6.sqlite [2/8] [=======================94% ] report6.sqlite [2/8] [=======================95% ] report6.sqlite [2/8] Line truncated
[3/8] Executing 'nvtx_sum' stats report
[4/8] Executing 'osrt_sum' stats report
Time (%) Total Time (ns) Num Calls Avg (ns) Med (ns) Min (ns) Max (ns) StdDev (ns) Name
-------- --------------- --------- ----------- ----------- -------- ---------- ----------- ----------------------
53.4 2194351837 32 68573494.9 100124295.0 1886 100179994 44439732.4 poll
42.9 1763348095 7 251906870.7 58081562.0 1467 1200215470 444651839.0 fread
2.9 119818701 600 199697.8 24235.0 1886 22988850 990232.3 ioctl
0.5 21655089 38 569870.8 7543.0 4819 20367773 3300181.2 fopen
0.1 2467848 25 98713.9 10057.0 7124 1710695 338493.4 mmap64
0.0 1330620 11 120965.5 45816.0 2375 925189 267079.5 write
0.0 871902 32 27246.9 3736.5 2374 383499 92136.2 fclose
0.0 634441 17 37320.1 9009.0 3353 404242 96596.2 mmap
0.0 389366 2 194683.0 194683.0 194648 194718 49.5 pthread_create
0.0 386078 55 7019.6 7193.0 3422 11105 1929.7 open64
0.0 135356 32 4229.9 1397.0 1396 91283 15885.4 fgets
0.0 107831 56 1925.6 1885.0 1397 12781 1497.9 fcntl
0.0 63625 2 31812.5 31812.5 1396 62229 43015.4 fflush
0.0 62369 9 6929.9 6635.0 4330 10127 1985.0 munmap
0.0 49378 6 8229.7 8171.5 4680 12502 2769.2 open
0.0 49098 34 1444.1 1431.5 1396 1886 85.5 pthread_mutex_trylock
0.0 41764 15 2784.3 2374.0 1886 5169 1140.3 read
0.0 23746 3 7915.3 8940.0 4819 9987 2732.1 pipe2
0.0 21929 2 10964.5 10964.5 8660 13269 3259.1 socket
0.0 11105 1 11105.0 11105.0 11105 11105 0.0 connect
0.0 10965 7 1566.4 1467.0 1396 1886 220.6 dup
0.0 7613 1 7613.0 7613.0 7613 7613 0.0 fopen64
0.0 6216 2 3108.0 3108.0 2933 3283 247.5 fwrite
0.0 4190 3 1396.7 1397.0 1396 1397 0.6 pthread_cond_broadcast
0.0 3841 1 3841.0 3841.0 3841 3841 0.0 bind
0.0 1885 1 1885.0 1885.0 1885 1885 0.0 listen
[5/8] Executing 'cuda_api_sum' stats report
Time (%) Total Time (ns) Num Calls Avg (ns) Med (ns) Min (ns) Max (ns) StdDev (ns) Name
-------- --------------- --------- ---------- ---------- -------- -------- ----------- ---------------------------------
73.6 59403870 3 19801290.0 22548780.0 21232 36833858 18559468.9 cudaMemcpy
14.8 11937011 12 994750.9 11280.0 1397 7337536 2110925.3 cudaFree
3.0 2437533 1 2437533.0 2437533.0 2437533 2437533 0.0 cudaFreeHost
2.4 1975324 1 1975324.0 1975324.0 1975324 1975324 0.0 cudaGetDeviceProperties_v2_v12000
1.7 1380415 1 1380415.0 1380415.0 1380415 1380415 0.0 cudaMallocHost
1.6 1294437 171 7569.8 4261.0 3702 400959 30590.6 cudaLaunchKernel
1.4 1159422 838 1383.6 1397.0 489 50775 1808.0 cuGetProcAddress_v2
1.2 974146 6 162357.7 144885.5 4749 353537 148396.1 cudaMalloc
0.1 52451 1 52451.0 52451.0 52451 52451 0.0 cudaMemset
0.0 36526 18 2029.2 1397.0 1327 7682 1573.8 cudaEventCreateWithFlags
0.0 34502 4 8625.5 7228.5 4400 15645 4884.7 cudaDeviceSynchronize
0.0 29613 18 1645.2 1466.0 978 4749 815.2 cudaEventDestroy
0.0 9919 4 2479.8 2375.0 2305 2864 258.3 cuInit
0.0 5238 3 1746.0 1397.0 978 2863 989.8 cuModuleGetLoadingMode
0.0 3283 2 1641.5 1641.5 1397 1886 345.8 cudaGetDriverEntryPoint_v11030
[6/8] Executing 'cuda_gpu_kern_sum' stats report
Time (%) Total Time (ns) Instances Avg (ns) Med (ns) Min (ns) Max (ns) StdDev (ns) Name
-------- --------------- --------- -------- -------- -------- -------- ----------- ----------------------------------------------------------------------------------------------------
67.6 10665692 49 217667.2 162496.0 77632 2419814 332250.3 matmul_forward_kernel(float *, const float *, const float *, const float *, int, int, int, int)
25.9 4086409 25 163456.4 163681.0 158656 166305 1957.5 layernorm_forward_kernel(float *, float *, float *, const float *, const float *, const float *, in…
1.4 226592 12 18882.7 18880.0 18720 19008 64.7 gelu_forward_kernel(float *, const float *, int)
1.4 215201 12 17933.4 17920.0 17856 18112 72.7 softmax_forward_kernel(float *, float, const float *, int, int)
0.9 145057 24 6044.0 6064.0 5664 6560 372.4 residual_forward_kernel(float *, float *, float *, int)
0.6 99136 12 8261.3 8224.0 8160 8512 105.5 attention_v_kernel(float *, float *, float *, int, int, int, int)
0.6 95616 12 7968.0 7968.0 7904 8032 36.1 preattention_scores_kernel(float *, float *, int, int, int, int, float *)
0.6 94912 12 7909.3 7888.0 7776 8128 121.2 permute_kernel(float *, float *, float *, const float *, int, int, int, int)
0.6 87072 1 87072.0 87072.0 87072 87072 0.0 encoder_forward_kernel(float *, const int *, const float *, const float *, int, int, int)
0.4 70528 12 5877.3 5872.0 5856 5920 24.9 unpermute_kernel(float *, float *, int, int, int, int)
[7/8] Executing 'cuda_gpu_mem_time_sum' stats report
Time (%) Total Time (ns) Count Avg (ns) Med (ns) Min (ns) Max (ns) StdDev (ns) Operation
-------- --------------- ----- ---------- ---------- -------- -------- ----------- ----------------------------
81.9 36688662 2 18344331.0 18344331.0 896 36687766 25941534.6 [CUDA memcpy Host-to-Device]
17.2 7687347 1 7687347.0 7687347.0 7687347 7687347 0.0 [CUDA memcpy Device-to-Host]
0.9 407265 1 407265.0 407265.0 407265 407265 0.0 [CUDA memset]
[8/8] Executing 'cuda_gpu_mem_size_sum' stats report
Total (MB) Count Avg (MB) Med (MB) Min (MB) Max (MB) StdDev (MB) Operation
---------- ----- -------- -------- -------- -------- ----------- ----------------------------
497.905 2 248.952 248.952 0.001 497.904 352.070 [CUDA memcpy Host-to-Device]
232.392 1 232.392 232.392 232.392 232.392 0.000 [CUDA memset]
51.511 1 51.511 51.511 51.511 51.511 0.000 [CUDA memcpy Device-to-Host]
Generated:
/u/asingh41/sp26_ece408_Temperature1.5/report6.nsys-rep
/u/asingh41/sp26_ece408_Temperature1.5/report6.sqlite