google-ai-edge ai-edge-torch issues

google-ai-edge / ai-edge-torch

Supporting PyTorch models with the Google AI Edge TFLite runtime.

Apache License 2.0

195 stars 21 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Converting Torch modules that use the max function

#81 hbellafkir opened 22 hours ago
0
Update README.md

#80 haozha111 closed 2 days ago
0
Not Able to run converted gemma model in tflite because of custom op odml.update_kv_cache

#79 akshatshah17 opened 2 days ago
2
[Prototype] Align SDPA definition with the gpu accelerator

#78 Linchenn opened 3 days ago
2
Fix torch requirement

#77 chunnienc closed 3 days ago
0
Relax torch version requirement

#76 chunnienc closed 3 days ago
0
Failed to convert gemma to tflite

#75 xujuntwt95329 opened 3 days ago
0
Cannot get ai-edge-tool running on WSL2 Ubuntu

#74 BmanClark opened 3 days ago
3
Revert "Change tokens from i64 to i32 (#68)"

#73 paulinesho closed 4 days ago
0
Update quantizer to new nightly

#72 paulinesho closed 2 days ago
1
Pin ai-edge-quantizer nightly version

#71 paulinesho closed 4 days ago
0
Update test_model_conversion.py

#70 haozha111 opened 1 week ago
0
Add a system overview of Generative API

#69 haozha111 closed 3 days ago
0
Change tokens from i64 to i32

#68 paulinesho closed 1 week ago
0
MacOS installation failed

#67 YuMinHui825 opened 1 week ago
5
Add channel last IO transformation API

#66 chunnienc closed 1 week ago
0
Interpreter can register builtin opcode after conversion

#65 kohheekyung opened 2 weeks ago
1
Fix BuildInterpolateCompositePass with aten decompositions

#64 chunnienc closed 1 week ago
0
Add README for unet layers.

#63 yichunk closed 2 weeks ago
0
Add fx pass to remove sdpa composite zero mask input

#62 chunnienc closed 2 weeks ago
0
Wrap embedding tables with HLFB

#61 paulinesho opened 2 weeks ago
0
Errors with incompatible versions of torch_xla and ai-edge-torch

#60 ccleavinger closed 2 weeks ago
2
Add search model api

#59 wuruoyu closed 2 weeks ago
0
Skip flaky quantization test temporarily

#58 paulinesho closed 2 weeks ago
0
Remove 5D tensor reshape in attention layer implementation.

#57 yichunk closed 3 weeks ago
3
support batch > 1 and add batch_size to model config.

#56 haozha111 closed 3 weeks ago
0
Add UNet model config and refactor diffusion model.

#55 yichunk closed 3 weeks ago
0
Update gemma and example model to use external KV.

#54 hheydary opened 3 weeks ago
0
LSTMs are being unrolled and decomposed after conversion

#53 Doomski99 opened 3 weeks ago
0
(DO NOT SUBMIT) Add source loc to debug info

#52 chunnienc opened 3 weeks ago
0
Fix broken decoder conversion of SD.

#51 yichunk closed 3 weeks ago
0
Add 2D blocks used in diffusion model of stable diffusion.

#50 yichunk closed 3 weeks ago
0
Add cross attention layer and refactor t5_attention

#49 yichunk closed 3 weeks ago
0
Fix activation_config renaming in autoencolder_loader

#48 yichunk closed 3 weeks ago
0
Add back missed deleted GELU_QUICK type.

#47 yichunk closed 3 weeks ago
0
Migrate quantization to use AI Edge Quantizer

#46 paulinesho closed 3 weeks ago
0
Update README.md

#45 haozha111 closed 3 weeks ago
0
Phi3 conversion OOM on A100

#44 a8nova opened 3 weeks ago
12
Text-To-Speech and Speech-To-Text models support

#43 bhanuprakashbv opened 3 weeks ago
0
Add GeGLU activation and change activation type to activation config

#42 yichunk closed 3 weeks ago
1
Add interpolate nearest composite support

#41 vamsimanchala closed 4 weeks ago
0
Update tf-nightly dep

#40 majiddadashi closed 4 weeks ago
0
Refactor decoder of stable diffusion

#39 yichunk closed 4 weeks ago
1
Bug: redefinition of symbol named 'gelu_decomp_0'

#38 Misaka17032 closed 2 weeks ago
1
update tf-nightly version so that models w/ external cache can be converted.

#37 haozha111 closed 4 weeks ago
0
External kv experiment

#36 hheydary closed 4 weeks ago
0
Migrate quantization to use AI Edge Quantizer

#35 paulinesho closed 4 weeks ago
0
Rename the gelu composite to avoid conflict

#34 majiddadashi closed 1 month ago
1
Enable min cut partitioner by default for layout transpose optimization

#33 chunnienc closed 3 weeks ago
0
Enable test for model conversion.

#32 haozha111 closed 1 month ago
1