pdufour commited on
Commit
a8fa7ff
1 Parent(s): b9ab9cc

Update EXPORT.md

Browse files
Files changed (1) hide show
  1. EXPORT.md +13 -8
EXPORT.md CHANGED
@@ -15,46 +15,51 @@ This will create an export of the onnx models.
15
 
16
  The following commands are all available:
17
 
18
- **General Commands**
19
 
20
  **all-in-one**
21
  Runs all steps (exporting, slimming, quantizing, cleaning, fixing GPU buffers) to produce fully prepared ONNX models.
22
 
23
- **Export Commands**
24
 
25
  **export**
 
26
  Combines export-abcd and export-e to generate ONNX models for all parts.
27
 
28
  **export-abcd**
 
29
  Exports model parts A, B, C, and D by running QwenVL_Export_ABCD.py.
30
 
31
  **export-e**
 
32
  Exports model part E by running QwenVL_Export_E.py.
33
 
34
- **Slimming Commands**
35
 
36
  ***slim**
 
37
  Reduces ONNX model size by removing unnecessary elements for optimized deployment.
38
 
39
- **Quantization Commands**
40
 
41
  **quantize**
 
42
  Quantizes all model parts (A, B, C, D, and E) to optimize size and performance.
43
 
44
  **quantize-%**
45
  Quantizes a specific model part (% can be A, B, C, D, or E) with targeted configurations.
46
 
47
- **Cleanup Commands**
48
 
49
  **clean-large-files**
50
  Deletes ONNX files larger than 2GB from the destination directory to retain appropriately sized models.
51
 
52
- **GPU Buffer Fix Command**
53
 
54
  **fix-gpu-buffers**
55
  Applies fixes to GPU buffers in ONNX files for part E to ensure GPU memory compatibility.
56
 
57
- **Combined Target**
58
 
59
- all
60
  Alias for all-in-one to run the full ONNX model preparation pipeline.
 
15
 
16
  The following commands are all available:
17
 
18
+ ## General Commands
19
 
20
  **all-in-one**
21
  Runs all steps (exporting, slimming, quantizing, cleaning, fixing GPU buffers) to produce fully prepared ONNX models.
22
 
23
+ ## Export Commands
24
 
25
  **export**
26
+
27
  Combines export-abcd and export-e to generate ONNX models for all parts.
28
 
29
  **export-abcd**
30
+
31
  Exports model parts A, B, C, and D by running QwenVL_Export_ABCD.py.
32
 
33
  **export-e**
34
+
35
  Exports model part E by running QwenVL_Export_E.py.
36
 
37
+ ## Slimming Commands
38
 
39
  ***slim**
40
+
41
  Reduces ONNX model size by removing unnecessary elements for optimized deployment.
42
 
43
+ ## Quantization Commands
44
 
45
  **quantize**
46
+
47
  Quantizes all model parts (A, B, C, D, and E) to optimize size and performance.
48
 
49
  **quantize-%**
50
  Quantizes a specific model part (% can be A, B, C, D, or E) with targeted configurations.
51
 
52
+ ## Cleanup Commands
53
 
54
  **clean-large-files**
55
  Deletes ONNX files larger than 2GB from the destination directory to retain appropriately sized models.
56
 
57
+ ## GPU Buffer Fix Command
58
 
59
  **fix-gpu-buffers**
60
  Applies fixes to GPU buffers in ONNX files for part E to ensure GPU memory compatibility.
61
 
62
+ ## Combined Target
63
 
64
+ **all**
65
  Alias for all-in-one to run the full ONNX model preparation pipeline.