v2.0.0-beta.3

[amcadmus]

New feature:

• derivatives for deep tensor (make deep tensor deriv again #805)

Performance improvement:

• speedup ROCm kernels which use atomicAdd (speedup ROCm kernel which uses atomicAdd #809 Synchronize CUDA _r modifications to ROCM #815 )
• speedup CUDA kernels (use atomicAdd inside) by reducing the global memory write (speedup CUDA kernels(use atomicAdd inside) by reducing the global memory write #811)

Enhancement:

• add type-embedding developer doc (add type-embedding developer doc #762)
• add model compression support for models with exclude_types feature (add model compression support for models with exclude_types feature #754)
• improve the doc and user interface of model compression (Add a more detailed introduction for model compression #772)
• allow c++ tests to run without internet (allow c++ tests to run without internet #785)
• support converting models generated in v1.3 to 2.0 compatibility (support converting models generated in v1.3 to 2.0 compatibility #725)
• give a default value to T and convert models from v1.2 to 2.0 compatibility (give a default value to T and convert models from v1.2 #789)
• improved documents for conda (add documents for conda #798)
• throw a message if tf runtime is incompatible (throw a message if tf runtime is incompatible #797)
• capture OOM and print debug message (add OOM message #801)

Bug fixings

• fix bug of custom op's multiple initialization (fix bug of custom op's multiple initialization #812)
• fix bug of empty input in gelu.cu (Fix bug of empty input in gelu.cu  #800)
• fix model compression bug of zero environment matrix (fix bug of empty environment matrix at model compression #808)
• box.npy is not necessary for nopbc (box.npy is not necessary for nopbc #810)
• fill rij with zero (fill rij with zero #818)
• NOT_LOADABLE should be tuple (NOT_LOADABLE should be tuple #825)

---

This discussion was created from the release v2.0.0-beta.3.