Google GitGoogle Git
Sign in
chromium/external/github.com/google-ai-edge/LiteRT/HEAD
  1. d67a796 Support DEQUANTIZE v5 constant weights and 6D DepthToSpace TRANSPOSE v6 in ML Drift TFLite parser. by Google AI Edge · 3 hours ago upstream/main
  2. fdc20d4 Verify FlatBuffer, subgraph, and tensor buffers directly in LiteRT. by Changming Sun CA · 5 hours ago
  3. a62f231 Update XNNPACK in XLA by Dillon Sharlet · 5 hours ago
  4. d3b84ea Pad model files with trailing zeros to avoid SIMD out-of-bounds reads by Terry Heo · 6 hours ago
  5. 717b5b9 Pad model files with trailing zeros to avoid SIMD out-of-bounds reads by Terry Heo · 6 hours ago
  6. f4b4155 Shard LiteRT fuzz tests and exclude builtin cases from XNNPACK targets. by Changming Sun CA · 6 hours ago
  7. f2843cf [Tensor API] ExecutionPlan lowering of the fused transformer ops, with conversion test. by Ping Yu · 6 hours ago
  8. 31af842 Exclude MoeExperts from ML Drift GPU and XNNPACK ATS test suites. by Gerardo Carranza · 7 hours ago
  9. 5bcbd4d Speed up externalize_tflite_flatbuffer for models with large embedded buffers. by Google AI Edge · 7 hours ago
  10. 92cee59 litert: Support internal host pointer I/O buffers to reduce delegate latency. by Google AI Edge · 8 hours ago
  11. faa57c6 [Tensor API] Qwen3.5: structured benchmark mode (CL6), perf smoke test, and Tensor API model bring-up agent skills. by Ping Yu · 8 hours ago
  12. 57d5dfb [Tensor API] Qwen3.5: GgufLoader for multi-format GGUF checkpoints and zero-copy GPU-resident KV caches / Gated DeltaNet states on ML Drift. by Ping Yu · 8 hours ago
  13. 6071e2c Discover `.dylib` shared libraries in the LiteRT dynamic loader on Apple platforms. by Andrew Zhang · 8 hours ago
  14. 3e49a37 Avoid O(N) list traversal in IrAllocator::EmplaceBack. by Doyoung Gwak · 9 hours ago
  15. dc7771a Internal change by Andrew Zhang · 9 hours ago
  16. 0e8e75a [Tensor API] Qwen3.5 text decoder skeleton: config, weight map, attention/MLP graph, GDN gap; ML Drift parity. by Ping Yu · 10 hours ago
  17. 8f9e43c Onboard qualcomm_ats_aot to the LiteRT ATS NPU AOT test suite. by Gerardo Carranza · 10 hours ago
  18. 29a5144 Create a YNNPACK backend for the Tensor API. by Quentin Khan · 11 hours ago
  19. afb64ec Materialize dynamic input dimensions to 1 in ExtraModel::MakeInputs. by Gerardo Carranza · 11 hours ago
  20. 7c5bdd9 Update XNNPACK and its dependencies (KleidiAI, slinky and cpuinfo). by Quentin Khan · 11 hours ago
  21. f49461f Remove uint8 special case in GetQuantizedConvolutionMultipler. by Esun Kim · 13 hours ago
  22. 585622e Unify in-place GPU state updates for `odml.short_conv_step` and `gated_delta_update` under `kGpuOptimizedInplace`. by Google AI Edge · 14 hours ago
  23. d45ae3c PR #127026: Fix L2 normalization fusion for multi-axis reductions by lao fan · 18 hours ago
  24. 4445554 Automated Code Change by Google AI Edge · 21 hours ago
  25. 0d58dc4 [Tensor API] Persist and round-trip hierarchical graph::Scope in TFLite FlatBuffers. by Ping Yu · 35 hours ago
  26. 3a26d12 Enable multi-iteration latency benchmarking on ATS ExtraModel test suites. by Gerardo Carranza · 3 days ago
  27. ab0f2b3 Add LayerNorm Generator to ATS. by Gerardo Carranza · 3 days ago
  28. 5d30579 Harden LiteRT model verification and shape inference. by Changming Sun CA · 3 days ago
  29. fdb3785 Internal only change by Weiyi Wang · 3 days ago
  30. d481ae5 Support STABLEHLO_COMPOSITE (odml.group_norm) with full reduction_axes [1, ..., n-1] when sub_type is not specified. by Google AI Edge · 3 days ago
  31. bfbe14d Expand `SLICE` fuzz test coverage in `slice_fuzz_test.cc`: by Changming Sun · 3 days ago
  32. a230575 Add MoeExperts (odml.moe_experts) Generator to ATS. by Gerardo Carranza · 3 days ago
  33. db2f08d Use top-level actions/cache in Linux container workflows by Terry Heo · 3 days ago
  34. 3d0b2bc Remove GitHub team link from LiteRT CODEOWNERS. by Weiyi Wang · 3 days ago
  35. 484131d Promote non-external `odml.cache_update` outputs to buffer storage in `LiteRtOpSelector`. by Gerardo Carranza · 3 days ago
  36. b0983dc Don't repeat KV heads in SDPA. by Quentin Khan · 4 days ago
  37. cdade5e Check mldrift_compatibility in RegisterGpuAccelerator on Android. by Alexandru Paiu · 4 days ago
  38. 481f1f6 Exclude Intel OpenVINO guitar validation tests from arm builds. by Matt Kreileder CA · 4 days ago
  39. 20c6a10 Allow the buffers in the `SafetensorLoader` to not be ordered by offset. by Quentin Khan · 4 days ago
  40. 4b5599a Merge pull request #10966 from google-ai-edge/litert_pr_1000636638 by jl45G · 4 days ago
  41. 4935c8f This is an internal change by Google AI Edge · 4 days ago
  42. 45b6211 This is an internal change by Google AI Edge · 4 days ago
  43. 1810bd3 This is an internal change by Google AI Edge · 4 days ago
  44. e8421a1 This is an internal change by Google AI Edge · 4 days ago
  45. 2b7daa1 This is an internal change by Google AI Edge · 4 days ago
  46. 469072e Drop IoT SHA256SUMS file. by Andrew Zhang · 4 days ago
  47. 8a0173a This is an internal change by Google AI Edge · 4 days ago
  48. b85895f This is an internal change by Google AI Edge · 4 days ago
  49. 99525ba This is an internal change by Google AI Edge · 4 days ago
  50. 9ea6c2f This is an internal change by Google AI Edge · 4 days ago
  51. 76f7e8b This is an internal change by Google AI Edge · 4 days ago
  52. c424f91 This is an internal change by Google AI Edge · 4 days ago
  53. 553876a This is an internal change by Google AI Edge · 4 days ago
  54. 24ace46 Ensure natural alignment for read-only tensor buffers in `Subgraph::SetTensorParametersReadOnly` by Gerardo Carranza · 4 days ago
  55. 0d64c92 Ensure natural alignment for read-only tensor buffers in `Subgraph::SetTensorParametersReadOnly` by Gerardo Carranza · 4 days ago
  56. 7a207625 Enforce 16-byte buffer alignment in FlatBuffer export. by Gerardo Carranza · 4 days ago
  57. e43a4e6 Enforce 16-byte buffer alignment in FlatBuffer export. by Gerardo Carranza · 4 days ago
  58. d6f5db3 Pass RuntimeBmm control tensor as a dynamic input for ATS. by Gerardo Carranza · 4 days ago
  59. 8ed4518 Support Bool, Int8, UInt8, and Int16 in untemplated SimpleBuffer::WriteRandom. by Gerardo Carranza · 4 days ago
  60. 9f02430 Add missing relu_min_max.cc source to QNN transformation CMake build. by Google AI Edge · 4 days ago
  61. 951b933 Fix QnnTypes.h include path in conv2d_test.cc by Aravind Murali · 4 days ago
  62. 6f0c911 Merge pull request #10742 from YingyingMaIntel:unflod_gqa_sequence by Copybara-Service · 4 days ago
  63. e0fd631 Merge pull request #9996 from graham0824:dev/hungjuiw/relu-min-max by Copybara-Service · 4 days ago
  64. 309cf08 Fold f32, DRQ, and Weight-Only Conv2D, DepthwiseConv2D, TransposeConv, and FC ops. by Google AI Edge · 4 days ago
  65. f8bccfc Merge pull request #10658 from akodanka:looped_write by Copybara-Service · 4 days ago
  66. d71ba31 Mark decomposition subgraphs as delegation-skippable in YNNPackDelegate. by Gerardo Carranza · 4 days ago
  67. 13daf66 Merge pull request #10629 from YingyingMaIntel:26b_moe_prefill_p2 by Copybara-Service · 4 days ago
  68. ea75322 Fold f32, DRQ, and Weight-Only Conv2D, DepthwiseConv2D, TransposeConv, and FC ops. by Artsiom Ablavatski · 4 days ago
  69. 8cf9a2f [tflite] Constant-fold PAD with a constant input. by Google AI Edge · 5 days ago
  70. 98737e7 Fix review comments by hungjuiw · 5 days ago
  71. 97dbc0a Refactor litert_api_no_jni to export sub-libraries. by Google AI Edge · 5 days ago
  72. 791e065 Use a `size_t` instead of an `int` in `LoadLastBuildStep`. by Quentin Khan · 5 days ago
  73. 86bbf78 Fix RoPE with src_sh.b != pos_sh.b by Google AI Edge · 5 days ago
  74. 29cf164 Qualcomm AI Engine Direct - Add FuseMinMaxOpsToReLUMinMax. by hungjuiw · 5 days ago
  75. 1ca66eb Refactor litert_api_no_jni to export sub-libraries. by Catalin Termure · 5 days ago
  76. b6bc769 PR #127955: Avoid re-logging non-variable input tensors post-evaluation in calibrator by RohithPariki · 5 days ago
  77. 97fbc68 PR #123123: security/fuzzing: add TFLite model load + interpreter invoke fuzz target by Layau Eulizier Jr · 5 days ago
  78. da71531 Add XNNPACK unit test for variable assignment with no outputs. by Google AI Edge · 5 days ago
  79. eba450b [tflite] Set hoist_constants_as_args=False in TFLiteJaxConverterV2 by George Necula · 5 days ago
  80. 77c7e83 Harden `SLICE` kernel and XNNPACK delegate validation and expand unit test coverage: by Changming Sun · 5 days ago
  81. 3cf7257 This is an internal change by Google AI Edge · 5 days ago
  82. 536d641 Harden convolution kernels and LiteRT model verification against int32_t overflows. by Changming Sun · 5 days ago
  83. 40e5f81 Harden convolution kernels and LiteRT model verification against int32_t overflows. by Changming Sun · 5 days ago
  84. 2e93f21 Automated Code Change by Google AI Edge · 5 days ago
  85. d15ef8f Fix float-to-double narrowing in GetQuantizedConvolutionMultipler for int8/int16. by Esun Kim · 5 days ago
  86. 8085eea Fix float-to-double narrowing in TFLite reference kernels. by Esun Kim · 5 days ago
  87. 5ef80b7 Fix float-to-double narrowing in TFLite quantized Mul kernel. by Esun Kim · 5 days ago
  88. 28c5c62 Add FuzzTest suites in `embedding_lookup_fuzz_test.cc` covering `EMBEDDING_LOOKUP` (dense, hybrid per-tensor/per-axis, and blockwise 2-bit/4-bit quantized) and `HASHTABLE_LOOKUP` (numeric and string tables), along with regression tests for sub-byte buffer sizing, blockwise scale overflow, extreme `int32_t` keys, and corrupted string offsets. by Changming Sun · 5 days ago
  89. bea3bc6 PR #10797: Qualcomm AI Engine Direct - Support 4-bit per-channel quantization. by Graham · 5 days ago
  90. 395f964 [Intel OV] Extend the MoE rewrite to multi-token chunks by yingyingma · 5 days ago
  91. 681fd6f [Intel OV] Rework WeightBank for MoE-generated constants by yingyingma · 5 days ago
  92. ef2055c Adopt LRTTensorType in LRTTensorBuffer and add signature/tensor introspection to LRTCompiledModel. by Fengwu Yao · 5 days ago
  93. 4d8b338 Replace legacy TFLite GPU delegate modules with ML Drift in LiteRT by Terry Heo · 5 days ago
  94. e8ea29d Add Qualcomm compiler graph transformations for Vision Encoder models. by Andrew Zhang · 5 days ago
  95. 9a0a72f Upload in-memory weights directly in LiteRtWeightLoader::UploadWeightsOnWebImpl by Google AI Edge · 5 days ago
  96. 01f7b0e Add a progress bar for model buffer serialization by Maria Lyubimtseva · 5 days ago
  97. 338d96e Harden lookup kernels and LiteRT model verification. by Changming Sun · 5 days ago
  98. 4b03036 Harden lookup kernels and LiteRT model verification. by Changming Sun · 5 days ago
  99. 62f1a9c Add acceleration status checks and accelerator discovery to LiteRT Rust bindings. by Google AI Edge · 5 days ago
  100. 62e92c8 Merge pull request #9216 from jatinwadhwa921:jatin_add_logs by Copybara-Service · 5 days ago