Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / zhihu/cubert issues and pull requests
#83 - 代码更改问题咨询
Issue -
State: open - Opened by zhengzhengwenbo over 1 year ago
#83 - 代码更改问题咨询
Issue -
State: open - Opened by zhengzhengwenbo over 1 year ago
#82 - 运行训练问题
Issue -
State: open - Opened by zhengzhengwenbo over 1 year ago
- 2 comments
#82 - 运行训练问题
Issue -
State: open - Opened by zhengzhengwenbo over 1 year ago
- 2 comments
#81 - Java api run slowerly than Python
Issue -
State: open - Opened by benlihua over 1 year ago
- 1 comment
#81 - Java api run slowerly than Python
Issue -
State: open - Opened by benlihua over 1 year ago
- 1 comment
#80 - what(): Unable to open vocab file
Issue -
State: open - Opened by liucong521 almost 2 years ago
#80 - what(): Unable to open vocab file
Issue -
State: open - Opened by liucong521 almost 2 years ago
#79 - 关于test中模型的其他输出
Issue -
State: closed - Opened by zlwzlwzlw almost 4 years ago
- 1 comment
#78 - 在cpu上的cubert推理速度比tf_bert慢
Issue -
State: closed - Opened by manutd12 almost 4 years ago
- 1 comment
#78 - 在cpu上的cubert推理速度比tf_bert慢
Issue -
State: closed - Opened by manutd12 almost 4 years ago
- 1 comment
#77 - 在进行build时执行cmake -DCMAKE_BUILD_TYPE=Release -DcuBERT_ENABLE_MKL_SUPPORT=ON .. make -j4,报错
Issue -
State: closed - Opened by manutd12 almost 4 years ago
- 3 comments
#77 - 在进行build时执行cmake -DCMAKE_BUILD_TYPE=Release -DcuBERT_ENABLE_MKL_SUPPORT=ON .. make -j4,报错
Issue -
State: closed - Opened by manutd12 almost 4 years ago
- 3 comments
#76 - solve infinite loop at bad utf8 string decoding
Pull Request -
State: closed - Opened by levyfan about 4 years ago
#76 - solve infinite loop at bad utf8 string decoding
Pull Request -
State: closed - Opened by levyfan about 4 years ago
#75 - infinite loop when handling bad utf8 string
Issue -
State: closed - Opened by oujiafan about 4 years ago
#75 - infinite loop when handling bad utf8 string
Issue -
State: closed - Opened by oujiafan about 4 years ago
#74 - 请问编译出来的libcuBert.so该如何使用?
Issue -
State: closed - Opened by yutaoxxx about 4 years ago
#74 - 请问编译出来的libcuBert.so该如何使用?
Issue -
State: closed - Opened by yutaoxxx about 4 years ago
#73 - Bump junit from 4.12 to 4.13.1 in /java
Pull Request -
State: closed - Opened by dependabot[bot] about 4 years ago
Labels: dependencies
#73 - Bump junit from 4.12 to 4.13.1 in /java
Pull Request -
State: closed - Opened by dependabot[bot] about 4 years ago
Labels: dependencies
#72 - fix cpu softmax nan on large input
Pull Request -
State: closed - Opened by levyfan about 4 years ago
#72 - fix cpu softmax nan on large input
Pull Request -
State: closed - Opened by levyfan about 4 years ago
#71 - transformer output nan in last few layer
Issue -
State: closed - Opened by 0rchard over 4 years ago
- 1 comment
#71 - transformer output nan in last few layer
Issue -
State: closed - Opened by 0rchard over 4 years ago
- 1 comment
#70 - 序列化任务, 跟官方bert运行结果不一致
Issue -
State: closed - Opened by ghost over 4 years ago
#70 - 序列化任务, 跟官方bert运行结果不一致
Issue -
State: closed - Opened by ghost over 4 years ago
#69 - could you provide the demo of how can generate input_ids input_mask、segment_ids by c++
Issue -
State: closed - Opened by ghost over 4 years ago
#69 - could you provide the demo of how can generate input_ids input_mask、segment_ids by c++
Issue -
State: closed - Opened by ghost over 4 years ago
#68 - CuBERT not utilizing all threads with multi-cpu
Issue -
State: open - Opened by volker42maru almost 5 years ago
- 4 comments
#68 - CuBERT not utilizing all threads with multi-cpu
Issue -
State: open - Opened by volker42maru almost 5 years ago
- 4 comments
#67 - support system installed mkl and protobuf
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#67 - support system installed mkl and protobuf
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#66 - merge query, key, value layer into 1 single strided gemm to improve perf
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
- 1 comment
#65 - use mkl_malloc and mkl_free to better performance reproducibility
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#65 - use mkl_malloc and mkl_free to better performance reproducibility
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#64 - further accelerate CPU gemm by mkl compact blas
Issue -
State: open - Opened by levyfan almost 5 years ago
Labels: enhancement
#64 - further accelerate CPU gemm by mkl compact blas
Issue -
State: open - Opened by levyfan almost 5 years ago
Labels: enhancement
#63 - upgrade to v0.0.5
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#62 - add python devel docker file
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#62 - add python devel docker file
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#61 - fix oom and more gpu info when oom
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#61 - fix oom and more gpu info when oom
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#60 - release version 0.0.4
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#60 - release version 0.0.4
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#59 - add cuBERT_get_gpu_count in API
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#59 - add cuBERT_get_gpu_count in API
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#58 - remove boxing in java and improve java speed
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#58 - remove boxing in java and improve java speed
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#57 - convert half to float in C++
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#56 - add dockerfile
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#56 - add dockerfile
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#55 - use external utf8proc
Pull Request -
State: closed - Opened by levyfan almost 5 years ago
#54 - 愿意来线下参加个分享吗?
Issue -
State: closed - Opened by linkinbird almost 5 years ago
#53 - build from source的时候遇到点问题
Issue -
State: closed - Opened by DamonCC about 5 years ago
- 10 comments
#53 - build from source的时候遇到点问题
Issue -
State: closed - Opened by DamonCC about 5 years ago
- 10 comments
#52 - out of memory when seq_length=384 and batch_size=1
Issue -
State: closed - Opened by onlyone509 about 5 years ago
- 2 comments
#52 - out of memory when seq_length=384 and batch_size=1
Issue -
State: closed - Opened by onlyone509 about 5 years ago
- 2 comments
#51 - undefined reference to `cublasGemmStridedBatchedEx'
Issue -
State: closed - Opened by onlyone509 about 5 years ago
- 3 comments
#51 - undefined reference to `cublasGemmStridedBatchedEx'
Issue -
State: closed - Opened by onlyone509 about 5 years ago
- 3 comments
#50 - 弱问怎么得到自己的bert_frozen_seq32.pb?
Issue -
State: closed - Opened by guotong1988 about 5 years ago
- 4 comments
#50 - 弱问怎么得到自己的bert_frozen_seq32.pb?
Issue -
State: closed - Opened by guotong1988 about 5 years ago
- 4 comments
#49 - java中测试mklBert计算时间很高
Issue -
State: closed - Opened by dawson-chen about 5 years ago
- 2 comments
#49 - java中测试mklBert计算时间很高
Issue -
State: closed - Opened by dawson-chen about 5 years ago
- 2 comments
#48 - mklBERT设置batch比较大的时候,性能很差,跟你的测试结果不一致。
Issue -
State: closed - Opened by comeby about 5 years ago
- 3 comments
#48 - mklBERT设置batch比较大的时候,性能很差,跟你的测试结果不一致。
Issue -
State: closed - Opened by comeby about 5 years ago
- 3 comments
#47 - 编译不通过
Issue -
State: closed - Opened by dawson-chen about 5 years ago
- 4 comments
#47 - 编译不通过
Issue -
State: closed - Opened by dawson-chen about 5 years ago
- 4 comments
#46 - upgrade version number to 0.0.3
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#46 - upgrade version number to 0.0.3
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#45 - switch to new API in python and java
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#44 - introduce new API for multiple model outputs at once
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#44 - introduce new API for multiple model outputs at once
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#43 - What's the setting of the benchmark?
Issue -
State: closed - Opened by leerumor over 5 years ago
- 4 comments
#43 - What's the setting of the benchmark?
Issue -
State: closed - Opened by leerumor over 5 years ago
- 4 comments
#42 - cuBERT运行环境要求
Issue -
State: open - Opened by Jackie127 over 5 years ago
- 2 comments
Labels: enhancement
#42 - cuBERT运行环境要求
Issue -
State: open - Opened by Jackie127 over 5 years ago
- 2 comments
Labels: enhancement
#41 - switch to protobuf-c to avoid conflicting with tensorflow itself
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#41 - switch to protobuf-c to avoid conflicting with tensorflow itself
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#40 - customize gemm algorithm for CUDA
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#40 - customize gemm algorithm for CUDA
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#39 - cuBERT 支持fine tune吗
Issue -
State: closed - Opened by starspringcloud over 5 years ago
- 1 comment
#39 - cuBERT 支持fine tune吗
Issue -
State: closed - Opened by starspringcloud over 5 years ago
- 1 comment
#38 - java单句子表征
Issue -
State: closed - Opened by Jackie127 over 5 years ago
- 1 comment
#38 - java单句子表征
Issue -
State: closed - Opened by Jackie127 over 5 years ago
- 1 comment
#37 - set D_GLIBCXX_USE_CXX11_ABI=0 and fix https://github.com/zhihu/cuBERT/issues/36
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#37 - set D_GLIBCXX_USE_CXX11_ABI=0 and fix https://github.com/zhihu/cuBERT/issues/36
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#36 - To be fully compatible with tensorflow, the library should be compiled with D_GLIBCXX_USE_CXX11_ABI=0
Issue -
State: closed - Opened by levyfan over 5 years ago
#36 - To be fully compatible with tensorflow, the library should be compiled with D_GLIBCXX_USE_CXX11_ABI=0
Issue -
State: closed - Opened by levyfan over 5 years ago
#35 - fix performance issue with blas_gemm half precision
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#35 - fix performance issue with blas_gemm half precision
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#34 - What mkl options have you used on benchmark?
Issue -
State: closed - Opened by hsh2438 over 5 years ago
- 1 comment
#34 - What mkl options have you used on benchmark?
Issue -
State: closed - Opened by hsh2438 over 5 years ago
- 1 comment
#33 - add java test in travis CI
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#33 - add java test in travis CI
Pull Request -
State: closed - Opened by levyfan over 5 years ago
#32 - java support error
Issue -
State: closed - Opened by wxrui over 5 years ago
- 7 comments
#32 - java support error
Issue -
State: closed - Opened by wxrui over 5 years ago
- 7 comments
#31 - Different result cmp to bert/extract_features.py
Issue -
State: closed - Opened by Damcy over 5 years ago
- 2 comments
#31 - Different result cmp to bert/extract_features.py
Issue -
State: closed - Opened by Damcy over 5 years ago
- 2 comments
#30 - compatible with CUDA<3.5
Pull Request -
State: closed - Opened by levyfan over 5 years ago