jina-ai / clip-as-service

πŸ„ Scalable embedding, reasoning, ranking for images and sentences with CLIP
https://clip-as-service.jina.ai
Other
12.48k stars 2.07k forks source link

fix: add models and md5 #783

Closed ZiniuYu closed 2 years ago

ZiniuYu commented 2 years ago

This pr adds open-clip model filenames and their corresponding md5 We now host a copy of open-clip models and convert them to onnx runtime

codecov[bot] commented 2 years ago

Codecov Report

Merging #783 (3d24db6) into main (7c8285b) will decrease coverage by 3.33%. The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #783      +/-   ##
==========================================
- Coverage   86.64%   83.30%   -3.34%     
==========================================
  Files          21       21              
  Lines        1108     1108              
==========================================
- Hits          960      923      -37     
- Misses        148      185      +37     
Flag Coverage Ξ”
cas 83.30% <100.00%> (-3.34%) :arrow_down:

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files Coverage Ξ”
server/clip_server/executors/clip_torch.py 87.93% <ΓΈ> (ΓΈ)
server/clip_server/model/clip_onnx.py 88.23% <ΓΈ> (ΓΈ)
server/clip_server/model/pretrained_models.py 84.12% <ΓΈ> (-1.59%) :arrow_down:
server/clip_server/model/openclip_model.py 78.57% <100.00%> (-10.72%) :arrow_down:
server/clip_server/model/trt_utils.py 56.04% <0.00%> (-27.48%) :arrow_down:
server/clip_server/model/clip_trt.py 71.05% <0.00%> (-21.06%) :arrow_down:

Continue to review full report at Codecov.

Legend - Click here to learn more Ξ” = absolute <relative> (impact), ΓΈ = not affected, ? = missing data Powered by Codecov. Last update 7c8285b...3d24db6. Read the comment docs.