nvidia
/

Qwen2.5-VL-7B-Instruct-FP8

Text Generation

Model Optimizer

Model card Files Files and versions

i-riyad commited on Oct 8

Commit

bbaf9a3

·

unverified ·

1 Parent(s): 033f56c

rope_scaling type fix

Signed-off-by: Riyad Islam <[email protected]>

Files changed (1) hide show

config.json +2 -2

config.json CHANGED Viewed

@@ -25,7 +25,7 @@
             24
         ],
         "rope_type": "default",
-        "type": "default"
     },
     "rope_theta": 1000000.0,
     "sliding_window": 32768,
@@ -86,7 +86,7 @@
                 24
             ],
             "rope_type": "default",
-            "type": "default"
         },
         "rope_theta": 1000000.0,
         "sliding_window": null,

             24
         ],
         "rope_type": "default",
+        "type": "mrope"
     },
     "rope_theta": 1000000.0,
     "sliding_window": 32768,
                 24
             ],
             "rope_type": "default",
+            "type": "mrope"
         },
         "rope_theta": 1000000.0,
         "sliding_window": null,