sarvex/DeepSpeed

main

DeepSpeed

Grade	Name	Complexity	Churn	Issues
F	deepspeed\ops\sparse_attention\matmul.py	117	3	209
F	deepspeed\runtime\pipe\module.py	116	3	88
F	deepspeed\runtime\data_pipeline\data_sampling\indexed_dataset.py	115	3	108
F	deepspeed\launcher\runner.py	105	5	63
F	deepspeed\ops\sparse_attention\sparsity_config.py	103	3	190
F	deepspeed\runtime\zero\partitioned_param_coordinator.py	103	4	60
F	deepspeed\utils\zero_to_fp32.py	102	8	127
F	deepspeed\runtime\hybrid_engine.py	101	4	95
F	deepspeed\runtime\swap_tensor\optimizer_utils.py	98	3	75
A	csrc\includes\conversion_utils.h	96	4	-
F	deepspeed\runtime\fp16\onebit\lamb.py	92	4	76
F	deepspeed\runtime\fp16\fused_optimizer.py	90	3	78
F	deepspeed\runtime\state_dict_factory.py	86	3	95
F	deepspeed\runtime\fp16\unfused_optimizer.py	84	3	55
A	csrc\transformer\ds_transformer_cuda.cpp	82	3	2
F	deepspeed\autotuning\scheduler.py	81	3	73
D	deepspeed\runtime\pipe\topology.py	80	3	42
F	deepspeed\compression\helper.py	79	3	58
D	deepspeed\launcher\multinode_runner.py	78	5	50
C	accelerator\cuda_accelerator.py	77	4	30