Learn sparse linear models
Project description
thinc: Learn super-sparse multi-class models
thinc is a Cython library for learning models with millions of parameters and dozens of classes. It drives https://spacy.io , a pipeline of very efficient NLP components. I’ve only used thinc from Cython; no real Python API is currently available.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
thinc-5.0.8.tar.gz
(690.8 kB
view hashes)
Built Distributions
thinc-5.0.8-cp35-none-win_amd64.whl
(387.1 kB
view hashes)
thinc-5.0.8-cp34-none-win_amd64.whl
(356.5 kB
view hashes)
thinc-5.0.8-cp27-none-win_amd64.whl
(361.4 kB
view hashes)
Close
Hashes for thinc-5.0.8-cp35-none-win_amd64.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | e95930a87061f2083998d45abf1e3c9b677da2423e04c0b031881c921d883909 |
|
MD5 | cd5d686a70a2ab57bd9c4dd2b1423611 |
|
BLAKE2b-256 | 441d0c6c9fba996dd604fe6058659d0b4c12b91df1486b23d3a986b5b0047049 |
Close
Hashes for thinc-5.0.8-cp35-cp35m-manylinux1_x86_64.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 503b550f48e908cff418e5967e3e005a9a6b247d75dcb7694d7b12ead3911be3 |
|
MD5 | 49fdf061b6d22a0cc0e35b7ae6af6a33 |
|
BLAKE2b-256 | 4bdc6b8c6fb1122d17c7ccd1dbdf5a0561503231a27044707d142c52faf7732c |
Close
Hashes for thinc-5.0.8-cp35-cp35m-macosx_10_6_intel.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 883d98ba21d969bc1ca4ce0ba09cd2b0070f65616ad812b5b2be8c4d13d066dc |
|
MD5 | 1ff212802060be9bd62780040371a456 |
|
BLAKE2b-256 | 79c20be717e3290d2df769189f09bebed03b201524ecc2e4167bda37611789de |
Close
Hashes for thinc-5.0.8-cp34-none-win_amd64.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 254774b165ff817b02a5a3eafde7774d7d94d88265433d169e022af47c7bc74c |
|
MD5 | e9ed48c14ae66d26efb146903a0febe8 |
|
BLAKE2b-256 | 6a73ee43355fa212f22e280523140179737a88d02901a217ff6c199607925d41 |
Close
Hashes for thinc-5.0.8-cp34-cp34m-manylinux1_x86_64.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | bd0993ba2f01ed023bc86e7ea3a860cd11872f1f5dca29bbf738518c333c38d3 |
|
MD5 | 3d9a83580cd66a6dafa83cb491a6787c |
|
BLAKE2b-256 | f8892db15dfb9213031de5e356cf32b225db122148596710c197e26eef515873 |
Close
Hashes for thinc-5.0.8-cp34-cp34m-macosx_10_6_intel.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 767aee351341b9166a32b869be4150ed4d2ab6d31390b47e3a1bfbe8a575892d |
|
MD5 | 629814e4a703541b4209a5930b305a16 |
|
BLAKE2b-256 | 7f3df68c416c3336973ada131aa8b6fb27793539c877bb3bfc797144d3274ee8 |
Close
Hashes for thinc-5.0.8-cp27-none-win_amd64.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 3bd4bff40d8c7bc021c517c463b94568f54874b8e265b72a9ac5c47d285b5962 |
|
MD5 | 71ad203ff8e24afb361b87b3b77b333c |
|
BLAKE2b-256 | 909afd2f442b045700d890e30fff43b346c25409878de6fd8d42a8c84008fd90 |
Close
Hashes for thinc-5.0.8-cp27-cp27mu-manylinux1_x86_64.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | fd0397b231102362c85cc40765be6a76a6c25b2faa8aabdb3ba43a9b29d5945d |
|
MD5 | fa3fc65ddd3f79f9a8d907463fe7e5ad |
|
BLAKE2b-256 | cfa752021ce2ece2c4101719d171134f707412982f79bffbf4df6d29d12a5069 |
Close
Hashes for thinc-5.0.8-cp27-cp27m-manylinux1_x86_64.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | af1ad7f841022fd9a0e3d84e9fe1a780276a99f9b71e0501dcb8a2093816ea9d |
|
MD5 | 217d545fc14f3dfeda5b2b42a58c289b |
|
BLAKE2b-256 | 692a015ae530c17d20b4c71334a0ca045f92a07670110796df6ea95c84ae2d36 |
Close
Hashes for thinc-5.0.8-cp27-cp27m-macosx_10_6_intel.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 43db9dccd6d456db8e453a22363e54352bab8bd6d3c3a479f0df43586d8bff74 |
|
MD5 | 83d21cb2e5c91c006e611eb4e4d74911 |
|
BLAKE2b-256 | 83f782db19ae13372d0f18b09bdcfce69db8ea4111f99ac24719a9764a2e3222 |