tflite-micro

vee1e/tflite-micro

Fork 0

mirror of https://github.com/vee1e/tflite-micro.git synced 2026-09-01 09:51:22 +00:00

Commit graph

33604e7e0c

Validate RESIZE_NEAREST_NEIGHBOR size tensor against output shape fix-resize-nn-size-validation Lakshit Verma 2026-08-26 04:39:38 +05:30
ebf99da4b2

docs(python): explain env vars in test runner (#3679) gh-readonly-queue/main/pr-3679-13cd6c1550f62718c134b03e8b0acd1f9f50b76e Ryan Kuester 2026-08-25 22:57:14 +00:00
2137ccab68 Sync from upstream TF. sync-from-upstream-tf TFLM-bot 2026-08-25 14:16:36 +00:00
13cd6c1550

feat(compression): reject empty compression spec (#3678) main Ryan Kuester 2026-08-24 23:11:41 +00:00
733736087d

test(compression): mix compressed and uncompressed inputs (#3675) Ryan Kuester 2026-08-24 15:54:55 +00:00
90b983c2b3

test(compression): run integration tests without compression flag (#3673) Ryan Kuester 2026-08-20 16:47:21 +00:00
6ea6dfc489

refactor(compression): compressors inherit from Compressor protocol (#3672) Ryan Kuester 2026-08-20 00:36:51 +00:00
27b78c9fd1

ci(pypi): file an issue on wheel build failure (#3669) Ryan Kuester 2026-08-19 20:42:47 +00:00
433055700d

test(compression): add proprietary model integration test (#3670) Ryan Kuester 2026-08-19 20:42:17 +00:00
f8c117b558

perf(pypi): build the wheel in opt mode (#3657) Ryan Kuester 2026-08-13 22:23:16 +00:00
5654901771

test(compression): add integration tests with TFLM interpreter (#3659) Ryan Kuester 2026-08-13 22:22:55 +00:00
f26158b1be

feat(compression): use DECODE operators in output models (#3654) Ryan Kuester 2026-08-12 10:42:41 -05:00
ff3209d088

Manually synced from upstream (#3650) Esun Kim 2026-08-12 08:25:39 -07:00
e6a8627f4e

fix(pypi): repair wheel build for bzlmod (#3653) Ryan Kuester 2026-08-11 09:53:45 -05:00
855170977b

Bump keras from 3.12.0 to 3.15.0 in /python dependabot/pip/python/keras-3.15.0 dependabot[bot] 2026-08-09 13:19:51 +00:00
763a8946b9

Bump cryptography from 46.0.3 to 50.0.0 in /python dependabot/pip/python/cryptography-50.0.0 dependabot[bot] 2026-08-04 18:56:28 +00:00
18b9e6f2a8

test(python): rewrite unsupported-compression test for legacy path (#3646) Ryan Kuester 2026-08-03 23:33:59 -05:00
b6c0c20762

Bump actions/stale from 10 to 11 dependabot/github_actions/actions/stale-11 dependabot[bot] 2026-07-30 01:13:21 +00:00
330b1747c9

feat(compression): add DECODE operator insertion (#3624) Ryan Kuester 2026-07-29 15:14:55 -05:00
b89fb3e06e

Tune hard_swish_test.cc (#3640) Esun Kim 2026-07-29 08:34:53 -07:00
d5d27c829c

Bump actions/setup-python from 6 to 7 dependabot/github_actions/actions/setup-python-7 dependabot[bot] 2026-07-23 01:13:22 +00:00
047c6914f9

Bump setuptools from 80.9.0 to 83.0.0 in /python dependabot/pip/python/setuptools-83.0.0 dependabot[bot] 2026-07-22 14:11:30 +00:00
8f1f3b2623

WHILE operator input/output copy fix (#3633) David Davis 2026-07-20 10:58:59 -07:00
0965635467

Fix benchmark tool alternate profiler (#3629) David Davis 2026-07-13 11:10:56 -07:00
35b20c5c74

Sync from upstream TF. (#3625) TFLM-bot 2026-07-13 11:31:17 -07:00
fddd3707a3

Sync from upstream TF. (#3617) TFLM-bot 2026-07-08 15:34:57 -07:00
5750228374

Add Reset support to CircularBuffer op (#3618) Esun Kim 2026-07-08 14:34:43 -07:00
bc0e8c658b

Re-enable 3D x 2D BatchMatMul dimension collapse for adj_x=true (#3599) Esun Kim 2026-07-08 08:32:05 -07:00
9664901486

feat(python): add alt decompression memory parameter to interpreter (#3614) Ryan Kuester 2026-07-06 11:38:39 -05:00
e142972d4f

Manual sync from sync from github.com/tensorflow/tensorflow (#3616) Esun Kim 2026-07-01 18:13:01 -07:00
fbda8b8d6b

Manual sync from sync from github.com/tensorflow/tensorflow (#3601) Esun Kim 2026-07-01 14:46:41 -07:00
ca0b8b1308

Added allow-unsafe-pr-checkout: true (#3615) Esun Kim 2026-07-01 13:59:22 -07:00
5fa112687b

ci: declare workflow-level contents: read on 3 workflows (#3558) Arpit Jain 2026-07-01 08:47:20 +09:00
6c67d478e3

feat(compression): add Huffman and Pruning compression support (#3612) Ryan Kuester 2026-06-30 17:53:59 -05:00
5f54b5be28

Bump actions/checkout from 6 to 7 (#3609) dependabot[bot] 2026-06-30 15:26:08 -07:00
aa7d6f9fd8

Added Error Handling & Defensive Programming Guide (#3539) Esun Kim 2026-06-30 15:24:50 -07:00
b5acfb0355

Bazel 8.7.0 (#3610) Esun Kim 2026-06-30 15:23:57 -07:00
348eed01b6

docs: update deprecated Silabs repository link (#3579) Aman Kumar 2026-06-26 05:21:59 +05:30
9bfaeadf48

feat(compression): add LUT compression plugin (#3608) Ryan Kuester 2026-06-25 01:44:48 +03:00
acc13e42eb

Update CMSIS-NN download (#3607) Måns Nilsson 2026-06-24 06:03:35 +02:00
074b75f8ec

feat(compression): add Compressor protocol (#3590) Ryan Kuester 2026-06-18 23:15:56 +03:00
7ca66d18e7

Ensure int64 accumulation for bias-less 16x8 ops in TFLM kernels (#3598) Esun Kim 2026-06-16 16:58:19 -07:00
b6ee3dc974

Fix missing quantizedBiasType setting for 16x8 requantization (#3597) Esun Kim 2026-06-16 08:53:23 -07:00
e0b2c281f2

feat(compression): add DECODE operator types and metadata (#3589) Ryan Kuester 2026-06-05 20:31:13 -05:00
eea46d3d61

chore(compression): remove test_models.py (#3587) Ryan Kuester 2026-06-04 13:44:14 -05:00
ac1fae3619

refactor(compression): replace test_models with model_editor in compress_test (#3586) Ryan Kuester 2026-06-02 11:09:54 -05:00
730449fef1

chore(compression): remove model_facade.py (#3585) Ryan Kuester 2026-06-01 18:03:06 -05:00
c2accf5c8f

refactor(compression): migrate compress.py from model_facade to model_editor (#3580) Ryan Kuester 2026-06-01 11:36:48 -05:00
63819227bc

Fix cmsis-nn pooling (#3584) Esun Kim 2026-06-01 09:24:20 -07:00
7c26381de5

feat(compression): implement model_editor for TFLite model manipulation (#3575) Ryan Kuester 2026-05-29 16:42:17 -05:00
db550c38c9

refactor(compression): hoist numpy dtype map into tensor_type (#3578) Ryan Kuester 2026-05-28 16:56:37 -05:00
cc99d0597f

Removed batch_size assert in depthwise_conv (#3562) Esun Kim 2026-05-27 09:41:22 -07:00
2ba41b0f3e

Fixed rank support in pad and min/max (#3563) Esun Kim 2026-05-27 09:41:17 -07:00
67fd13064a

Format (#3577) Esun Kim 2026-05-27 06:39:06 -07:00
7914a2d7a1

Clean-up unused bzl load (#3576) Esun Kim 2026-05-26 15:55:55 -07:00
068c6a59b6

build(bazel): inject tflite_micro shim via tflm_py_* wrappers (#3573) Ryan Kuester 2026-05-26 14:51:58 -05:00
d0890bb456

Updated CMSIS_NN to the latest (6d9d61d) (#3565) Esun Kim 2026-05-21 15:05:29 -07:00
ac2c346736

Fix (#3567) Esun Kim 2026-05-21 14:53:37 -07:00
a9952453ab

Improve Check TfLite Files (#3566) Esun Kim 2026-05-21 14:40:16 -07:00
e0bece470e

Exclude //tensorflow/lite/micro/examples/mnist_lstm:train from (m|a|ub)san builds gonnet-patch-1 Pedro Gonnet 2026-05-08 12:59:56 +02:00
9f5ac257ee

No tensorflow 2 (#3548) Esun Kim 2026-05-06 14:17:46 -07:00
1c2bae841e

No tensorflow 1 (#3546) Esun Kim 2026-05-06 11:31:12 -07:00
cf8c42a16f

Align Conv INT16 accumulator logic with FullyConnected (#3537) Esun Kim 2026-05-06 06:02:51 -07:00
51bee03bed

Default to 64-bit accumulation for 16x8 Fully Connected without bias (#3522) Esun Kim 2026-04-09 11:35:03 -07:00
c36885e9e6

[Fix] Batch MatMul Op (#3413) Esun Kim 2026-04-09 11:34:59 -07:00
f97216c244

Sync from upstream TF. (#3488) TFLM-bot 2026-03-31 07:00:31 -07:00
f5302ed4fa

Add 16bit xtensa depthwise conv kernel support (#3481) Nicolás Arrieta Larraza 2026-03-18 17:23:13 +01:00
f2b2b3f51c

Sync from upstream TF. (#3461) TFLM-bot 2026-02-12 14:18:36 -08:00
399e48b8a9

[CI] Clean up bzl (#3478) Esun Kim 2026-02-12 14:05:54 -08:00
5aa7a2e4b4

Format (#3477) Esun Kim 2026-02-12 13:49:57 -08:00
257ca0f8d0

Removed //pythons from windows/bazel (#3476) Esun Kim 2026-02-12 13:36:39 -08:00
5b1f1007d0

Simplfied (#3475) Esun Kim 2026-02-12 13:25:47 -08:00
13cabf14a0

One more fix for Windows Bazel (#3474) Esun Kim 2026-02-12 13:06:24 -08:00
49126dfa28

Another windows fix (#3473) Esun Kim 2026-02-12 11:36:06 -08:00
0349d05e45

Hmm (#3472) Esun Kim 2026-02-12 11:19:58 -08:00
626c337c8e

Fixed again (#3471) Esun Kim 2026-02-12 11:04:57 -08:00
ba95ae6972

Fixed generate_per_layer_tests.py (#3470) Esun Kim 2026-02-12 10:58:01 -08:00
9759b03620

Fix the test (#3469) Esun Kim 2026-02-12 10:56:21 -08:00
3c27730240

Ignore non-zero exit code for bazel query in windows tests (#3468) Esun Kim 2026-02-12 10:39:34 -08:00
c981fbabc4 Exempt TFLM-bot and dependabot from BUG= check (#3467) Esun Kim 2026-02-12 10:30:30 -08:00
088df4ca7f

Let's be picky on Windows Bazel (#3465) Esun Kim 2026-02-12 10:20:47 -08:00
1e8997af71

[CI] Cache for Bazel on Linux (#3464) Esun Kim 2026-02-12 10:06:52 -08:00
a6d3820832

+build:windows_ci --build_python_zip=false (#3462) Esun Kim 2026-02-12 09:33:55 -08:00
6fd703818f

Added MODULE.bazel.lock (#3460) Esun Kim 2026-02-12 09:08:58 -08:00
e16b58697d

More windows (#3459) Esun Kim 2026-02-12 08:44:10 -08:00
df1992c680

Windows tmp (#3457) Esun Kim 2026-02-11 21:41:11 -08:00
ca24d40fc7

[CI] Windows .bazelrc (#3456) Esun Kim 2026-02-11 20:52:38 -08:00
3210fee460

symlinks (#3454) Esun Kim 2026-02-11 20:06:01 -08:00
8ae46c56fb

[CI] Windows CI (#3453) Esun Kim 2026-02-11 18:01:27 -08:00
eb51847198

Test bazel cache with hosted (#3452) Esun Kim 2026-02-11 16:26:11 -08:00
aa40ee75fb

CI has 1 hour timeout (#3451) Esun Kim 2026-02-11 16:07:28 -08:00
b4380d3128

Bump wheel from 0.45.1 to 0.46.2 in /python dependabot/pip/python/wheel-0.46.2 dependabot[bot] 2026-02-11 22:12:45 +00:00
807c35d6a9

Bazel Module (#3364) Esun Kim 2026-02-11 14:11:47 -08:00
183da3e302

Typo in code (#3447) Esun Kim 2026-02-11 12:34:42 -08:00
d7f46b6fb7

[Doc] Fix typos (#3446) Esun Kim 2026-02-11 09:18:49 -08:00
510a6c505c

pick 42ff343b # Fix handling of empty CSV files in generate_cc_arrays.py (#3423) Esun Kim 2026-02-10 15:25:19 -08:00
d87fa35a95

Bump peter-evans/create-pull-request from 8.0.0 to 8.1.0 (#3416) dependabot[bot] 2026-02-10 15:24:44 -08:00
58181a4261

Bump actions/github-script from 7 to 8 (#3431) dependabot[bot] 2026-02-10 15:22:55 -08:00
0e6f7b32a0

More (#3445) Esun Kim 2026-02-10 15:19:03 -08:00
a553df6a50

test_makefile on linux-x86-n2-16 (#3444) Esun Kim 2026-02-10 15:14:25 -08:00