hpcaitech / PaLM-colossalai

Scalable PaLM implementation of PyTorch

hpcaitech/PaLM-colossalai Issues

torch.distributed.elastic.multipro cessing.errors.ChildFailedError
Updated a year ago
ModuleNotFoundError: No module named 'torch._six'
Updated a year ago
Can I run this on one rtx 4070 ti?
Updated a year ago
There should be a version mismatching problem
Closed 2 years ago4
Fails with cannot import colo_set_process_memory_fraction in Docker
Updated 2 years ago
Warning When Using Different HuggingFace Datasets
Closed 2 years ago2
Gemini badcase
Closed 2 years ago5
Gemin+2.5D badcase
Updated 2 years ago
bash ./tools/download_token.py </PATH/TO/TOKENIZER/>
Closed 2 years ago
[feature] add model checkpointing
Updated 2 years ago
[feature] Add performance and scalability results
Updated 2 years ago
GitHub encourages communication and ongoing review.
Closed 2 years ago2
Have you really reproduced PaLM or just joking?
Closed 2 years ago2