Update inference-cache-config/qwen2.5-large.json
Browse files
inference-cache-config/qwen2.5-large.json
CHANGED
|
@@ -1,5 +1,19 @@
|
|
| 1 |
{
|
| 2 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
{
|
| 4 |
"batch_size": 1,
|
| 5 |
"sequence_length": 4096,
|
|
|
|
| 1 |
{
|
| 2 |
+
"deepseek-ai/DeepSeek-R1-Distill-Qwen-32B": [
|
| 3 |
+
{
|
| 4 |
+
"batch_size": 1,
|
| 5 |
+
"sequence_length": 4096,
|
| 6 |
+
"num_cores": 8,
|
| 7 |
+
"auto_cast_type": "bf16"
|
| 8 |
+
},
|
| 9 |
+
{
|
| 10 |
+
"batch_size": 8,
|
| 11 |
+
"sequence_length": 4096,
|
| 12 |
+
"num_cores": 8,
|
| 13 |
+
"auto_cast_type": "bf16"
|
| 14 |
+
}
|
| 15 |
+
],
|
| 16 |
+
"Qwen/Qwen2.5-32B-Instruct": [
|
| 17 |
{
|
| 18 |
"batch_size": 1,
|
| 19 |
"sequence_length": 4096,
|