Skip to content

Commit 99ba48f

Browse files
committed
Merge branch 'grpo-latest-rebase-main' of https://github.com/hpcaitech/ColossalAI into grpo-latest-rebase-main
2 parents 762150c + 94e972f commit 99ba48f

File tree

4 files changed

+12
-4
lines changed

4 files changed

+12
-4
lines changed

.github/workflows/run_chatgpt_examples.yml

Lines changed: 5 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -21,7 +21,7 @@ jobs:
2121
container:
2222
image: image-cloud.luchentech.com/hpcaitech/pytorch-cuda:2.2.2-12.1.0
2323
options: --gpus all --rm -v /data/scratch/examples-data:/data/scratch/examples-data --shm-size=10.24gb
24-
timeout-minutes: 60
24+
timeout-minutes: 180
2525
defaults:
2626
run:
2727
shell: bash
@@ -34,6 +34,10 @@ jobs:
3434
pip install --no-cache-dir -v -e .
3535
3636
- name: Install ChatGPT
37+
env:
38+
CFLAGS: "-O1"
39+
CXXFLAGS: "-O1"
40+
MAX_JOBS: 4
3741
run: |
3842
pip install flash-attn --no-build-isolation
3943
cd applications/ColossalChat

.github/workflows/run_chatgpt_unit_tests.yml

Lines changed: 5 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -21,7 +21,7 @@ jobs:
2121
container:
2222
image: image-cloud.luchentech.com/hpcaitech/pytorch-cuda:2.2.2-12.1.0
2323
options: --gpus all --rm -v /data/scratch/examples-data:/data/scratch/examples-data
24-
timeout-minutes: 30
24+
timeout-minutes: 180
2525
defaults:
2626
run:
2727
shell: bash
@@ -30,6 +30,10 @@ jobs:
3030
uses: actions/checkout@v2
3131

3232
- name: Install ChatGPT
33+
env:
34+
CFLAGS: "-O1"
35+
CXXFLAGS: "-O1"
36+
MAX_JOBS: 4
3337
run: |
3438
pip install flash-attn --no-build-isolation
3539
cd applications/ColossalChat

applications/ColossalChat/coati/distributed/grpo_consumer.py

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -530,4 +530,4 @@ def state_dict(self):
530530
model = self.policy_model.unwrap()
531531
state_dict = model.state_dict()
532532
state_dict["consumer_global_step"] = torch.tensor([self.global_step], device=self.device)
533-
return state_dict
533+
return state_dict

colossalai/shardformer/modeling/qwen3.py

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -273,7 +273,7 @@ def qwen3_for_causal_lm_forward(
273273
hidden_states: Optional[torch.FloatTensor] = None,
274274
stage_index: Optional[List[int]] = None,
275275
shard_config: ShardConfig = None,
276-
**kwargs
276+
**kwargs,
277277
):
278278
r"""
279279
Args:

0 commit comments

Comments
 (0)