Julien Simon commited on
Commit
58285ac
1 Parent(s): ce1f985

Add SuperNova AWQ benchmarks

Browse files
Files changed (1) hide show
  1. results_arcee_supernova.py +16 -0
results_arcee_supernova.py CHANGED
@@ -12,6 +12,22 @@ results_arcee_supernova = {
12
  "tokensPerSecond": "33",
13
  "notes": "MAX_INPUT_TOKENS: 8192, MAX_TOTAL_TOKENS: 16384",
14
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
15
  {
16
  "instanceType": "inf2.24xlarge",
17
  "configurations": [
 
12
  "tokensPerSecond": "33",
13
  "notes": "MAX_INPUT_TOKENS: 8192, MAX_TOTAL_TOKENS: 16384",
14
  },
15
+ {
16
+ "instanceType": "p4d.24xlarge",
17
+ "quantization": "awq",
18
+ "container": "TGI 2.2.0",
19
+ "status": "OK",
20
+ "tokensPerSecond": "58",
21
+ "notes": "MAX_INPUT_TOKENS: 16384, MAX_TOTAL_TOKENS: 32768",
22
+ },
23
+ {
24
+ "instanceType": "p5.48xlarge",
25
+ "quantization": "awq",
26
+ "container": "TGI 2.2.0",
27
+ "status": "OK",
28
+ "tokensPerSecond": "???",
29
+ "notes": "MAX_INPUT_TOKENS: 16384, MAX_TOTAL_TOKENS: 32768",
30
+ },
31
  {
32
  "instanceType": "inf2.24xlarge",
33
  "configurations": [