core
|
up
|
2023-09-26 14:30:47 +08:00 |
data
|
up
|
2023-09-26 14:30:47 +08:00 |
fp16_deprecated
|
change megatron to ascendspeed
|
2023-06-10 21:26:01 +08:00 |
model
|
baichuan13B模型
|
2023-09-28 10:30:20 +08:00 |
mpu
|
up
|
2023-09-26 14:30:47 +08:00 |
ops
|
Won't Import ascendspeed.ops if unneccssary
|
2023-08-16 09:35:10 +08:00 |
optimizer
|
up
|
2023-09-26 14:30:47 +08:00 |
tokenizer
|
up
|
2023-09-26 14:30:47 +08:00 |
__init__.py
|
update ascendspeed/__init__.py.
|
2023-09-26 08:56:56 +00:00 |
arguments.py
|
up
|
2023-09-26 14:30:47 +08:00 |
checkpointing.py
|
up
|
2023-09-26 14:30:47 +08:00 |
data_classes.py
|
up
|
2023-09-26 14:30:47 +08:00 |
enums.py
|
change megatron to ascendspeed
|
2023-06-10 21:26:01 +08:00 |
error_utils.py
|
up
|
2023-09-26 14:30:47 +08:00 |
global_vars.py
|
up
|
2023-09-26 14:30:47 +08:00 |
indexer.py
|
create megatron core
|
2023-07-24 15:00:57 +08:00 |
initialize.py
|
up
|
2023-09-26 14:30:47 +08:00 |
learning_rates.py
|
change megatron to ascendspeed
|
2023-06-10 21:26:01 +08:00 |
memory.py
|
change megatron to ascendspeed
|
2023-06-10 21:26:01 +08:00 |
microbatches.py
|
change megatron to ascendspeed
|
2023-06-10 21:26:01 +08:00 |
package_info.py
|
change megatron to ascendspeed
|
2023-06-10 21:26:01 +08:00 |
text_generation_utils.py
|
up
|
2023-09-26 14:30:47 +08:00 |
timers.py
|
add ffts mode
|
2023-07-25 14:14:28 +08:00 |
training.py
|
up
|
2023-09-26 14:30:47 +08:00 |
utils.py
|
up
|
2023-09-26 14:30:47 +08:00 |