Skip to content
Draft
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 8 additions & 0 deletions autoparallel/tools/overlap_simulator/colls32_8.table
Original file line number Diff line number Diff line change
@@ -0,0 +1,8 @@
Group Group Size Collective 1MB (ms) 2MB (ms) 4MB (ms) 8MB (ms) 16MB (ms) 32MB (ms) 64MB (ms) 128MB (ms) 256MB (ms) 512MB (ms) 1024MB (ms) 2048MB (ms)
------- ------------ -------------------------- ---------- ---------- ---------- ---------- ----------- ----------- ----------- ------------ ------------ ------------ ------------- -------------
1 8 all_gather_into_tensor 0.0495 0.0716 0.1138 0.1953 0.3584 0.6846 1.3371 2.642 5.2518 10.4714 20.9105 41.7888
1 8 reduce_scatter_tensor 0.0173 0.0238 0.0368 0.0495 0.0716 0.1138 0.1953 0.3584 0.6846 1.3371 2.642 5.2518
1 8 all_reduce 0.028 0.041 0.0628 0.0849 0.1292 0.2179 0.3822 0.7084 1.3609 2.6658 5.2756 10.4952
0 32 all_gather_into_tensor 1.0136 1.7497 3.1512 5.86 11.2777 22.113 43.7835 87.1247 173.807 347.171 693.901 1387.36
0 32 reduce_scatter_tensor 0.2114 0.2612 0.3608 0.4615 0.6455 1.0136 1.7497 3.1512 5.86 11.2777 22.113 43.7835
0 32 all_gather_into_tensor_out 1.0136 1.7497 3.1512 5.86 11.2777 22.113 43.7835 87.1247 173.807 347.171 693.901 1387.36
7 changes: 7 additions & 0 deletions autoparallel/tools/overlap_simulator/colls8_8.table
Original file line number Diff line number Diff line change
@@ -0,0 +1,7 @@
Group Group Size Collective 1MB (ms) 2MB (ms) 4MB (ms) 8MB (ms) 16MB (ms) 32MB (ms) 64MB (ms) 128MB (ms) 256MB (ms) 512MB (ms) 1024MB (ms) 2048MB (ms)
------- ------------ -------------------------- ---------- ---------- ---------- ---------- ----------- ----------- ----------- ------------ ------------ ------------ ------------- -------------
1 8 all_reduce 0.028 0.041 0.0628 0.0849 0.1292 0.2179 0.3822 0.7084 1.3609 2.6658 5.2756 10.4952
1 8 all_gather_into_tensor 0.0495 0.0716 0.1138 0.1953 0.3584 0.6846 1.3371 2.642 5.2518 10.4714 20.9105 41.7888
0 8 reduce_scatter_tensor 0.0866 0.1151 0.1566 0.2397 0.4059 0.7181 1.3297 2.5531 4.9998 9.8931 19.6798 39.2532
0 8 all_gather_into_tensor_out 0.2397 0.4059 0.7181 1.3297 2.5531 4.9998 9.8931 19.6798 39.2532 78.4001 156.694 313.281
0 8 all_gather_into_tensor 0.2397 0.4059 0.7181 1.3297 2.5531 4.9998 9.8931 19.6798 39.2532 78.4001 156.694 313.281
8,954 changes: 8,954 additions & 0 deletions autoparallel/tools/overlap_simulator/repro_llama3_8b_bw_256_1d_32layers.py

Large diffs are not rendered by default.

11,446 changes: 11,446 additions & 0 deletions autoparallel/tools/overlap_simulator/repro_llama3_8b_bw_256_2d_32layers.py

Large diffs are not rendered by default.

8,953 changes: 8,953 additions & 0 deletions autoparallel/tools/overlap_simulator/repro_llama3_8b_bw_64_1d_32layers.py

Large diffs are not rendered by default.

5,783 changes: 5,783 additions & 0 deletions autoparallel/tools/overlap_simulator/repro_llama3_8b_bw_64_2d_32layers.py

Large diffs are not rendered by default.

4,153 changes: 4,153 additions & 0 deletions autoparallel/tools/overlap_simulator/repro_llama3_8b_fw_256_1d_32layers.py

Large diffs are not rendered by default.

5,658 changes: 5,658 additions & 0 deletions autoparallel/tools/overlap_simulator/repro_llama3_8b_fw_256_2d_32layers.py

Large diffs are not rendered by default.

4,153 changes: 4,153 additions & 0 deletions autoparallel/tools/overlap_simulator/repro_llama3_8b_fw_64_1d_32layers.py

Large diffs are not rendered by default.

5,657 changes: 5,657 additions & 0 deletions autoparallel/tools/overlap_simulator/repro_llama3_8b_fw_64_2d_32layers.py

Large diffs are not rendered by default.

Loading
Loading