github/ColossalAI
1
0
Fork 0
mirror of https://github.com/hpcaitech/ColossalAI.git synced 2025-10-29 04:40:36 +00:00
Code Issues Packages Projects Releases Wiki Activity

[legacy] clean up legacy code (#4743)

Browse Source 
* [legacy] remove outdated codes of pipeline (#4692)

* [legacy] remove cli of benchmark and update optim (#4690)

* [legacy] remove cli of benchmark and update optim

* [doc] fix cli doc test

* [legacy] fix engine clip grad norm

* [legacy] remove outdated colo tensor (#4694)

* [legacy] remove outdated colo tensor

* [test] fix test import

* [legacy] move outdated zero to legacy (#4696)

* [legacy] clean up utils (#4700)

* [legacy] clean up utils

* [example] update examples

* [legacy] clean up amp

* [legacy] fix amp module

* [legacy] clean up gpc (#4742)

* [legacy] clean up context

* [legacy] clean core, constants and global vars

* [legacy] refactor initialize

* [example] fix examples ci

* [example] fix examples ci

* [legacy] fix tests

* [example] fix gpt example

* [example] fix examples ci

* [devops] fix ci installation

* [example] fix examples ci
This commit is contained in:
Hongxin Liu
2023-09-18 16:31:06 +08:00
committed by GitHub
parent 32e7f99416
commit b5f9e37c70
342 changed files with 2919 additions and 4182 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
examples/tutorial/sequence_parallel/loss_func/cross_entropy.py
View File

@@ -1,7 +1,8 @@
from colossalai.context.parallel_mode import ParallelMode
import torch
from torch.cuda.amp import custom_bwd, custom_fwd
from colossalai.legacy.context.parallel_mode import ParallelMode
class _VocabCrossEntropy(torch.autograd.Function):
@@ -24,8 +25,7 @@ class _VocabCrossEntropy(torch.autograd.Function):
# [*, partition-vocab-size] and target to a 1-D tensor of size [*].
logits_2d = vocab_parallel_logits.view(-1, vocab_parallel_logits.size(-1))
masked_target_1d = masked_target.view(-1)
arange_1d = torch.arange(start=0, end=logits_2d.size()[0],
device=logits_2d.device)
arange_1d = torch.arange(start=0, end=logits_2d.size()[0], device=logits_2d.device)
predicted_logits_1d = logits_2d[arange_1d, masked_target_1d]
predicted_logits_1d = predicted_logits_1d.clone().contiguous()
predicted_logits = predicted_logits_1d.view_as(target)
@@ -58,10 +58,8 @@ class _VocabCrossEntropy(torch.autograd.Function):
grad_2d = grad_input.view(-1, partition_vocab_size)
# Add the gradient from matching classes.
arange_1d = torch.arange(start=0, end=grad_2d.size()[0],
device=grad_2d.device)
grad_2d[arange_1d, masked_target_1d] -= (
1.0 - target_mask.view(-1).float())
arange_1d = torch.arange(start=0, end=grad_2d.size()[0], device=grad_2d.device)
grad_2d[arange_1d, masked_target_1d] -= (1.0 - target_mask.view(-1).float())
# Finally elementwise multiplication with the output gradients.
grad_input.mul_(grad_output.unsqueeze(dim=-1))