Commits · Oysiyl/AI-QR-code-generator

udate pizza example

151a131

Oysiyl commited on 4 days ago

Fix CUDA initialization error: call compile_models_with_aoti() inside main block after ZeroGPU init

c883ade

Oysiyl commited on 5 days ago

Switch torch.compile mode from 'reduce-overhead' to 'max-autotune' for faster inference

3490b07

Oysiyl commited on 5 days ago

Add .detach() before all .cpu().numpy() calls to fix gradient error with torch.compile

1652ae0

Oysiyl commited on 5 days ago

Fix all warmup parameters: border_size=4, error_correction='Medium (15%)', module_size=12, module_drawer='Square'

4eba716

Oysiyl commited on 5 days ago

Fix warmup parameters: use module_size=12 and error_correction='Medium (15%)' to match generate_all_triton_kernels.py

89a59b3

Oysiyl commited on 5 days ago

Fix module_drawer parameter: use 'Square' instead of 'square'

72ccbb3

Oysiyl commited on 5 days ago

Fix AOTI: use correct imports (aoti_capture, aoti_compile, aoti_apply) with torch.compile fallback

f205cc5

Oysiyl commited on 5 days ago

Simplify AOT compilation: use aoti_compile_model API instead of full pipeline

b09e131

Oysiyl commited on 5 days ago

Fix AOT compilation: use qr_text parameter instead of text_input

eee3359

Oysiyl commited on 5 days ago

Fix AOT compilation: move call after pipeline function definitions

134881e

Oysiyl commited on 5 days ago

Add AOT compilation for ZeroGPU cold start optimization

32d227e

Oysiyl Claude Sonnet 4.5 commited on 5 days ago

Increase ZeroGPU duration to 720s for local testing

fdbb0c1

Oysiyl Claude Sonnet 4.5 commited on 6 days ago

Increase ZeroGPU duration to 120s and disable share

0bd4d46

Oysiyl Claude Sonnet 4.5 commited on 6 days ago

Enable torch.compile for 1.5-1.7× speedup

576412f

Oysiyl Claude Sonnet 4.5 commited on 6 days ago

fix torch dynamo issue in freeu

413ecd6

Oysiyl commited on 8 days ago

increase time limit from 60 to 75

bc85fe6

Oysiyl commited on 8 days ago

add archive unpacking plus enable torch.compile skip if mps

0a9ecd7

Oysiyl commited on 8 days ago

disable vaetiling looks not so good

6dcf22b

Oysiyl commited on 9 days ago

disable torch.compile for now

a25a409

Oysiyl commited on 9 days ago

enable torch,compile with fixes

96f8c12

Oysiyl commited on 9 days ago

fixing issues

1f16a19

Oysiyl commited on 9 days ago

disable compilation for now

8eb6a33

Oysiyl commited on 9 days ago

torch.compile, bfloat16

0d712f8

Oysiyl commited on 9 days ago

Remove progress tracking, add essential fixes only

00847d8

Oysiyl commited on 9 days ago

memory improvements mostly

f7291d8

Oysiyl commited on 9 days ago

adjust description, tabs reordering, remove text baes examples from artistic tab

58a4bdb

Oysiyl commited on 12 days ago

add gallery, better exampels, add example with neapolitano pizza

faa27af

Oysiyl commited on 12 days ago

fix bug with wrong counter on steps; controlnet sliders, updating examples for artistic pipeline

b738f5e

Oysiyl commited on 12 days ago

possible to run whole thing on Mac Silicon

67c7b1f

Oysiyl commited on 12 days ago

disable cache examples

edb13a9

Oysiyl commited on 17 days ago

improvements to QR code generation, shareable JSON

e173e7b

Oysiyl commited on 17 days ago

updating params

1318560

Oysiyl commited on 17 days ago

add more params in advanced settings

515964d

Oysiyl commited on 22 days ago

added a new border plus default values for freeU

6587064

Oysiyl commited on 30 days ago

add FreeU

5b5b5fe

Oysiyl commited on Nov 21

wi-fi example requires more pixels

3351453

Oysiyl commited on Nov 21

add second version in a new tab

18e10c2

Oysiyl commited on Nov 20

set caching examples to false

5ea37e4

Oysiyl commited on Nov 15

show intermediate steps of a pipeline

f9eb2b6

Oysiyl commited on Nov 15

more lines to show error message and update error message

ee35bee

Oysiyl commited on Nov 15

soft exception handling

4b1e9b9

Oysiyl commited on Nov 15

change max_size for calendar event example

24a7976
verified

Oysiyl commited on Nov 15

catch RuntimeError as well

bcb5600

Oysiyl commited on Nov 15

increaze image size for example with VCARD

ed12b85

Oysiyl commited on Nov 15

mre specific exception

b2d89b4

Oysiyl commited on Nov 15

multiple changes

7a24597

Oysiyl commited on Nov 15

Update app.py

b7f734a
verified

Oysiyl commited on Nov 13

Update app.py

4bc2b89
verified

Oysiyl commited on Nov 13

Update app.py (#6)

cee611a
verified

Oysiyl

huggerhuggy commited on Nov 13

Commit History

udate pizza example 151a131

Fix CUDA initialization error: call compile_models_with_aoti() inside main block after ZeroGPU init c883ade

Switch torch.compile mode from 'reduce-overhead' to 'max-autotune' for faster inference 3490b07

Add .detach() before all .cpu().numpy() calls to fix gradient error with torch.compile 1652ae0

Fix all warmup parameters: border_size=4, error_correction='Medium (15%)', module_size=12, module_drawer='Square' 4eba716

Fix warmup parameters: use module_size=12 and error_correction='Medium (15%)' to match generate_all_triton_kernels.py 89a59b3

Fix module_drawer parameter: use 'Square' instead of 'square' 72ccbb3

Fix AOTI: use correct imports (aoti_capture, aoti_compile, aoti_apply) with torch.compile fallback f205cc5

Simplify AOT compilation: use aoti_compile_model API instead of full pipeline b09e131

Fix AOT compilation: use qr_text parameter instead of text_input eee3359

Fix AOT compilation: move call after pipeline function definitions 134881e

Add AOT compilation for ZeroGPU cold start optimization 32d227e

Increase ZeroGPU duration to 720s for local testing fdbb0c1

Increase ZeroGPU duration to 120s and disable share 0bd4d46

Enable torch.compile for 1.5-1.7× speedup 576412f

fix torch dynamo issue in freeu 413ecd6

increase time limit from 60 to 75 bc85fe6

add archive unpacking plus enable torch.compile skip if mps 0a9ecd7

disable vaetiling looks not so good 6dcf22b

disable torch.compile for now a25a409

enable torch,compile with fixes 96f8c12

fixing issues 1f16a19

disable compilation for now 8eb6a33

torch.compile, bfloat16 0d712f8

Remove progress tracking, add essential fixes only 00847d8

memory improvements mostly f7291d8

adjust description, tabs reordering, remove text baes examples from artistic tab 58a4bdb

add gallery, better exampels, add example with neapolitano pizza faa27af

fix bug with wrong counter on steps; controlnet sliders, updating examples for artistic pipeline b738f5e

possible to run whole thing on Mac Silicon 67c7b1f

disable cache examples edb13a9

improvements to QR code generation, shareable JSON e173e7b

updating params 1318560

add more params in advanced settings 515964d

added a new border plus default values for freeU 6587064

add FreeU 5b5b5fe

wi-fi example requires more pixels 3351453

add second version in a new tab 18e10c2

set caching examples to false 5ea37e4

show intermediate steps of a pipeline f9eb2b6

more lines to show error message and update error message ee35bee

soft exception handling 4b1e9b9

change max_size for calendar event example 24a7976 verified

catch RuntimeError as well bcb5600

increaze image size for example with VCARD ed12b85

mre specific exception b2d89b4

multiple changes 7a24597

Update app.py b7f734a verified

Update app.py 4bc2b89 verified

Update app.py (#6) cee611a verified

udate pizza example

151a131

Fix CUDA initialization error: call compile_models_with_aoti() inside main block after ZeroGPU init

c883ade

Switch torch.compile mode from 'reduce-overhead' to 'max-autotune' for faster inference

3490b07

Add .detach() before all .cpu().numpy() calls to fix gradient error with torch.compile

1652ae0

Fix all warmup parameters: border_size=4, error_correction='Medium (15%)', module_size=12, module_drawer='Square'

4eba716

Fix warmup parameters: use module_size=12 and error_correction='Medium (15%)' to match generate_all_triton_kernels.py

89a59b3

Fix module_drawer parameter: use 'Square' instead of 'square'

72ccbb3

Fix AOTI: use correct imports (aoti_capture, aoti_compile, aoti_apply) with torch.compile fallback

f205cc5

Simplify AOT compilation: use aoti_compile_model API instead of full pipeline

b09e131

Fix AOT compilation: use qr_text parameter instead of text_input

eee3359

Fix AOT compilation: move call after pipeline function definitions

134881e

Add AOT compilation for ZeroGPU cold start optimization

32d227e

Increase ZeroGPU duration to 720s for local testing

fdbb0c1

Increase ZeroGPU duration to 120s and disable share

0bd4d46

Enable torch.compile for 1.5-1.7× speedup

576412f

fix torch dynamo issue in freeu

413ecd6

increase time limit from 60 to 75

bc85fe6

add archive unpacking plus enable torch.compile skip if mps

0a9ecd7

disable vaetiling looks not so good

6dcf22b

disable torch.compile for now

a25a409

enable torch,compile with fixes

96f8c12

fixing issues

1f16a19

disable compilation for now

8eb6a33

torch.compile, bfloat16

0d712f8

Remove progress tracking, add essential fixes only

00847d8

memory improvements mostly

f7291d8

adjust description, tabs reordering, remove text baes examples from artistic tab

58a4bdb

add gallery, better exampels, add example with neapolitano pizza

faa27af

fix bug with wrong counter on steps; controlnet sliders, updating examples for artistic pipeline

b738f5e

possible to run whole thing on Mac Silicon

67c7b1f

disable cache examples

edb13a9

improvements to QR code generation, shareable JSON

e173e7b

updating params

1318560

add more params in advanced settings

515964d

added a new border plus default values for freeU

6587064

add FreeU

5b5b5fe

wi-fi example requires more pixels

3351453

add second version in a new tab

18e10c2

set caching examples to false

5ea37e4

show intermediate steps of a pipeline

f9eb2b6

more lines to show error message and update error message

ee35bee

soft exception handling

4b1e9b9

change max_size for calendar event example

24a7976
verified

catch RuntimeError as well

bcb5600

increaze image size for example with VCARD

ed12b85

mre specific exception

b2d89b4

multiple changes

7a24597

Update app.py

b7f734a
verified

Update app.py

4bc2b89
verified

Update app.py (#6)

cee611a
verified