OpenEmbedding is a distributed framework to accelerate TensorFlow training and support TensorFlow Serving. It uses the parameter server architecture to store the Embedding Layer. So that single machine memory is not the limit of model size. OpenEmbedding can cooperate with all-reduce framework to support both data parallel and model parallel.
1 maintainer
Metadata
1 maintainer
Release files for openembedding 0.1.2
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| openembedding-0.1.2.tar.gz | 21.8 MB | Details |
Release files / openembedding-0.1.2.tar.gz
| Download URL | openembedding-0.1.2.tar.gz |
|---|---|
| Size | 21.8 MB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
15aae49db02e65db99857ab62ebeeb34d8874bbc4982aee1f07c2fbcd222c1b0
|
|
BLAKE2b-256 checksum How to use checksums |
c7c633dd9305228b99e66b8a28be6952cd0b0ae7cc2fb295fd66fecbf0181af9
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/3.5.0 importlib_metadata/4.8.2 pkginfo/1.7.1 requests/2.22.0 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.8.10
|
