Add more gpt2 configurations
Browse files
inference-cache-config/gpt2.json
CHANGED
@@ -3,13 +3,19 @@
|
|
3 |
{
|
4 |
"batch_size": 1,
|
5 |
"sequence_length": 1024,
|
6 |
-
"num_cores":
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
"auto_cast_type": "fp16"
|
8 |
},
|
9 |
{
|
10 |
"batch_size": 16,
|
11 |
"sequence_length": 1024,
|
12 |
-
"num_cores":
|
13 |
"auto_cast_type": "fp16"
|
14 |
}
|
15 |
]
|
|
|
3 |
{
|
4 |
"batch_size": 1,
|
5 |
"sequence_length": 1024,
|
6 |
+
"num_cores": 2,
|
7 |
+
"auto_cast_type": "fp16"
|
8 |
+
},
|
9 |
+
{
|
10 |
+
"batch_size": 4,
|
11 |
+
"sequence_length": 1024,
|
12 |
+
"num_cores": 2,
|
13 |
"auto_cast_type": "fp16"
|
14 |
},
|
15 |
{
|
16 |
"batch_size": 16,
|
17 |
"sequence_length": 1024,
|
18 |
+
"num_cores": 2,
|
19 |
"auto_cast_type": "fp16"
|
20 |
}
|
21 |
]
|