Making large AI models cheaper, faster and more accessible
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
Zangwei Zheng d1e5bafcd4 [NFC] polish colossalai/auto_parallel/tensor_shard/deprecated/__init__.py code style (#2291) 2 years ago
..
_C [optimizer] add div_scale for optimizers (#2117) 2 years ago
amp [NFC] polish colossalai/amp/naive_amp/grad_scaler/dynamic_grad_scaler.py code style (#2299) 2 years ago
auto_parallel [NFC] polish colossalai/auto_parallel/tensor_shard/deprecated/__init__.py code style (#2291) 2 years ago
builder [NFC] polish colossalai/builder/__init__.py code style (#1560) 2 years ago
cli [NFC] polish colossalai/cli/benchmark/benchmark.py code style (#2287) 2 years ago
communication improved allgather & reducescatter for 3d 2 years ago
context [hotfix] Fixing the bug related to ipv6 support 2 years ago
device [device] update flatten device mesh usage (#2079) 2 years ago
engine [engin/schedule] use p2p_v2 to recontruct pipeline_schedule (#1408) 2 years ago
fx [auto-parallel] refactoring ColoTracer (#2118) 2 years ago
gemini [Gemini] fix the convert_to_torch_module bug (#2269) 2 years ago
kernel [builder] MOE builder (#2277) 2 years ago
logging [logger] hotfix, missing _FORMAT (#2231) 2 years ago
nn [builder] MOE builder (#2277) 2 years ago
pipeline [example] add benchmark (#2276) 2 years ago
registry
tensor [autoparallel] fix runtime apply memory estimation (#2281) 2 years ago
testing [amp] add gradient clipping for unit tests (#2283) 2 years ago
trainer [polish] remove useless file _mem_tracer_hook.py (#1963) 2 years ago
utils [builder] unified cpu_optim fused_optim inferface (#2190) 2 years ago
zero [zero] polish low level zero optimizer (#2275) 2 years ago
__init__.py [setup] supported conda-installed torch (#2048) 2 years ago
constants.py updated tp layers 2 years ago
core.py
global_variables.py updated tp layers 2 years ago
initialize.py