-
Building `mlir` with `-DLLVM_ENABLE_WARNINGS=OFF` results in an error that `-no-warn-on-unused-template-args` is not defined for internal tablegen executable `mlir-src-sharder`.
```bash
cmake -G "…
-
Currently writes and reads can be executed on any shard but later writes need anyway to bounce to shard 0 due to raft group0 requirement. Perhaps it makes sense to improve cache (containing full auth …
-
Implement with support for passivation.
- should keep track of non properly passivated actors and restart them on the next start
- passivate after a configurable timeout
-
I expected this to work like `MLIR_PDLL_TABLEGEN` or `MLIR_TABLEGEN` but it doesn't. Building with `MLIR_SRC_SHARDER_TABLEGEN` set still causes a nested build to start, presumably to build `mlir-src-s…
-
-
When building from source, shaders aren't visible to greetd-mini-wl-greeter, so executing it from anywhere other than the directory including the shaders leads to a crash with the following error:
…
-
> ./toolkits/collaborative_filtering/itemcf --distance=4 --training=dataset --K=350 --quiet=0 --nshards=1 --clean_cache=1 --min_allowed_intersection=48 --allow_zeros=1
> WARNING: common.hpp(print_cop…
-
### 🚀 The feature, motivation and pitch
# 🚀 Feature
Provide a detailed API design for high-level PyTorch Sharding architecture. These generic sharding interfaces are for PyTorch users to shard m…
-
From `TODO`:
> union storage target. sharder can be thought of a specialization of union. sharder already unions, but has a hard-coded policy of where to put new blobs. union could a library (used by…
-
operations/secure_random/secure_random.cc:1:10: fatal error: tensorflow/core/util/work_sharder.h: No such file or directory
1 | #include "tensorflow/core/util/work_sharder.h"
| ^~…