Commit Graph

  • a4f0fc96c0 Integrate user sharding hints and HLO sharding annotations across MLIR dialects and ZML core, and remove the now‑unused module options arguments. Tarry Singh 2023-03-21 10:50:39 +0000
  • e30e35deeb Update benchmark example to use new user sharding hints and drop deprecated module options. Foke Singh 2023-03-20 15:31:44 +0000
  • 8746a5ce78 Expose zml/test_runner.zig publicly to enable users to employ the async test runner. Made the dependency on zml explicit and suggest treating test_runner as a zig_library rather than a filegroup. Tarry Singh 2023-03-16 13:22:35 +0000
  • fe531aef06 Clarify HuggingFace token handling in workspace, noting the standard CLI location and adding support for an environment variable. Foke Singh 2023-03-14 15:28:03 +0000
  • cd2f2209d0 Create token directory if it does not exist. Foke Singh 2023-03-13 15:31:13 +0000
  • 70d40208a2 runtimes/cuda: Fix version variable definitions in the build script to enable successful CUDA builds. Tarry Singh 2023-03-09 11:31:02 +0000
  • 7ef67eea27 zml: Relocate tests next to the functions they verify and remove obsolete dynamicSlice1d test. Tarry Singh 2023-03-08 14:10:11 +0000
  • dfa71018a5 zml: Remove pjrtx wrapper, migrate remaining helpers to their native modules, and fix blocking issue in Event.await. Tarry Singh 2023-03-06 17:05:56 +0000
  • 0c126c2e12 runtimes/cuda: Upgrade CUDA to 12.6.2 and cuDNN to 9.4.0. Tarry Singh 2023-03-03 15:17:26 +0000
  • f595d22134 runtimes/rocm: Upgrade ROCm to version 6.2.2. Tarry Singh 2023-03-01 13:15:50 +0000
  • ecf52ad724 zml.tokenizer: Implement proper byte fallback support by converting hex byte strings (e.g., “<0x40>”) to their characters and splitting unknown UTF‑8 codepoints into bytes, fixing tokenization. Tarry Singh 2023-02-28 14:40:25 +0000
  • 2f129f76c9 Add in-process sharding support across core ZML components (platform, shape, tensor, MLIR generation, buffers, and PJRT integration) Tarry Singh 2023-02-24 17:33:14 +0000
  • cad1a688da Add sharding usage to the benchmark and simple_layer example programs. Foke Singh 2023-02-23 11:18:27 +0000
  • fc718ab649 Add StableHLO bindings for versioning functions, enabling portable serialization of StableHLO. Tarry Singh 2023-02-22 15:41:33 +0000
  • 8fa3878fc3 PJRT: Add handling for rank‑0 case in getDimensions to avoid null pointer usage when num_dims is zero. Tarry Singh 2023-02-17 10:47:15 +0000
  • 639f5cd994 Replace log with select for generating the attention mask to avoid NaNs on zero values. Tarry Singh 2023-02-16 10:36:23 +0000
  • 24a7c98476 Implement scatterSlices functionality. Tarry Singh 2023-02-14 13:52:49 +0000
  • 934acb35a8 zml: initialize Tensor.min and Tensor.max reductions with proper extreme values to ensure correct results Tarry Singh 2023-02-10 12:28:41 +0000
  • be6328813d zml: clean up dead and commented code; note that copyslice is currently broken and pending reimplementation Tarry Singh 2023-02-08 17:13:47 +0000
  • 058e1415fa zml: deprecate buggy Tensor.chunk; introduce chunkExact and chunkAllowTrailing with clarified behavior Tarry Singh 2023-02-07 12:42:34 +0000
  • 7e131a106b Update examples/MODULE.bazel.lock to reflect XLA version bump. Foke Singh 2023-02-03 14:13:21 +0000
  • 0606ea1d7c Update Bazel workspace and runtime BUILD files to newer XLA, StableHLO, and LLVM versions, enabling batching‑dims support for the gather operator. Tarry Singh 2023-02-01 15:58:30 +0000
  • 897786e440 aio: correct refAllDecls handling for yaml and nemo modules Tarry Singh 2023-01-31 11:58:58 +0000
  • 7dcd8b516c zml/nn: fix resize implementations (resizeBilinear and resizeBicubic) and expand refAllDecl usage; all tests pass Tarry Singh 2023-01-27 14:35:11 +0000
  • 5e1688cbfd aio: refactor PyTorch model parsing for better readability and optimize slice handling Tarry Singh 2023-01-25 12:16:27 +0000
  • ebdb8db213 zml/tests: re‑enable all Zig tests, fix precision issue by switching to f32, and add refAllDecls to ensure all declarations are tested Tarry Singh 2023-01-23 16:28:19 +0000
  • f39b16e13d zml/test_runner: add optional filtering of test functions via command‑line argument, allowing selective execution of tests (e.g., bazel run //zml:test -- sdpa) Tarry Singh 2023-01-20 13:50:36 +0000
  • b961856e5f zml/tensor: correct typo in uniform comment ('substract' → 'subtract') Tarry Singh 2023-01-19 12:20:40 +0000
  • ccdf218961 Add multi‑axis, batched gatherValues support to tensor, shape, nn, quantization, and torch modules. Tarry Singh 2023-01-18 12:03:48 +0000
  • 16e066ec69 Add llama example demonstrating the new gatherValues functionality. Foke Singh 2023-01-11 09:58:09 +0000
  • 48b671f100 Fix CollectionOver scope error in ActivationCollector and clean dead code/comments in zml_utils.py Tarry Singh 2023-01-10 09:43:03 +0000
  • 04ad137417 Update howto_torch2zml docs to explain why the output variable can be None. Foke Singh 2023-01-09 17:05:09 +0000
  • fab1c93d5b docs: first model – fix const/var bug and enforce 80‑column width Foke Singh 2023-01-06 10:34:44 +0000
  • eded305649 Add initial documentation and example projects for ZML, covering how‑to guides, tutorials, and benchmark examples. Foke Singh 2023-01-03 10:21:07 +0000
  • 266da6d4be Add initial Bazel build configuration, async runtime implementation, and core MLIR dialect definitions for ZML. Tarry Singh 2023-01-02 14:28:25 +0000