Skip to content

Commit 5baeadb

Browse files
authored
misc: deployment meta data for llama (#299)
solutions_index.yaml is a file necessary for generating a deployment configuration by using CfgGen2. The file is added.
1 parent 2b0017a commit 5baeadb

File tree

1 file changed

+50
-0
lines changed

1 file changed

+50
-0
lines changed
Lines changed: 50 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,50 @@
1+
3:
2+
- batch_size: 2
3+
template_name: batch2_gpus3.json
4+
throughput: 6.878416069177055
5+
total_latency: 763.4156208798601
6+
4:
7+
- batch_size: 2
8+
template_name: batch2_gpus4.json
9+
throughput: 9.870116833658855
10+
total_latency: 768.5197830131934
11+
5:
12+
- batch_size: 2
13+
template_name: batch2_gpus5.json
14+
throughput: 11.360584453032693
15+
total_latency: 819.5614043465267
16+
6:
17+
- batch_size: 2
18+
template_name: batch2_gpus6.json
19+
throughput: 14.463189225066683
20+
total_latency: 819.5614043465268
21+
7:
22+
- batch_size: 2
23+
template_name: batch2_gpus7.json
24+
throughput: 16.26837492034193
25+
total_latency: 844.6573759981395
26+
8:
27+
- batch_size: 2
28+
template_name: batch2_gpus8.json
29+
throughput: 19.172022089452042
30+
total_latency: 846.5416830951349
31+
9:
32+
- batch_size: 2
33+
template_name: batch2_gpus9.json
34+
throughput: 21.529689936552092
35+
total_latency: 860.3947014131934
36+
10:
37+
- batch_size: 2
38+
template_name: batch2_gpus10.json
39+
throughput: 23.05239047851694
40+
total_latency: 892.7109025274995
41+
11:
42+
- batch_size: 2
43+
template_name: batch2_gpus11.json
44+
throughput: 25.6446473051836
45+
total_latency: 880.6010983411635
46+
12:
47+
- batch_size: 2
48+
template_name: batch2_gpus12.json
49+
throughput: 27.603873718139077
50+
total_latency: 889.3276958391205

0 commit comments

Comments
 (0)