AXERA-TECH
/

Qwen3-VL-8B-Instruct

Image-Text-to-Text

Qwen3-VL-2B-Instruct

Qwen3-VL-4B-Instruct

Qwen3-VL-8B-Instruct

Model card Files Files and versions

lihongjie007 commited on Nov 11

Commit

f3c9f23

·

verified ·

1 Parent(s): 415d3c7

Update README.md

Files changed (1) hide show

README.md +2 -2

README.md CHANGED Viewed

@@ -47,12 +47,12 @@ For those who are interested in model conversion, you can try to export axmodel
 **Image Process**
 |Chips| input size | image num | image encoder | ttft(168 tokens) | w8a16 | CMM | Flash |
 |--|--|--|--|--|--|--|--|
-|AX650| 384*384 | 1 | 280 ms | 1476 ms | 2.5 tokens/sec| GiB | 14 GiB |
 **Video Process**
 |Chips| input size | image num | image encoder |ttft(600 tokens) | w8a16 | CMM | Flash |
 |--|--|--|--|--|--|--|--|
-|AX650| 384*384 | 8  | 1114 ms | 4520 ms | 2.5 tokens/sec| GiB | 14 GiB |
 The DDR capacity refers to the CMM memory that needs to be consumed. Ensure that the CMM memory allocation on the development board is greater than this value.

 **Image Process**
 |Chips| input size | image num | image encoder | ttft(168 tokens) | w8a16 | CMM | Flash |
 |--|--|--|--|--|--|--|--|
+|AX650| 384*384 | 1 | 280 ms | 1476 ms | 2.5 tokens/sec| 11.8 GB | 14 GB |
 **Video Process**
 |Chips| input size | image num | image encoder |ttft(600 tokens) | w8a16 | CMM | Flash |
 |--|--|--|--|--|--|--|--|
+|AX650| 384*384 | 8  | 1114 ms | 4520 ms | 2.5 tokens/sec| 11.8 GB | 14 GB |
 The DDR capacity refers to the CMM memory that needs to be consumed. Ensure that the CMM memory allocation on the development board is greater than this value.