Skip to content

Support agent training - #7

Merged
zhuzilin merged 2 commits into
THUDM:mainfrom
yitianlian:agent_train
Jun 24, 2025
Merged

Support agent training#7
zhuzilin merged 2 commits into
THUDM:mainfrom
yitianlian:agent_train

Conversation

@yitianlian

Copy link
Copy Markdown
Collaborator

No description provided.

Comment thread docs/en/agent_training.md
# Convert model
PYTHONPATH=/root/Megatron-LM python tools/convert_hf_to_torch_dist.py \
--hf-checkpoint /root/hf_models/deepseek-ai--DeepSeek-R1-Distill-Qwen-7B \
--save /root/megatron_model/DeepSeek-R1-Distill-Qwen-7B-25.02

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please rename the ckpt the same as other docs. Also, we no longer need -25.02 suffix.

@zhuzilin
zhuzilin merged commit 5a1393c into THUDM:main Jun 24, 2025
rysaya pushed a commit to rysaya/slime that referenced this pull request Aug 15, 2025
m1nttee pushed a commit to m1nttee/slime that referenced this pull request Nov 24, 2025
update verl-quick-start in verl
PopSoda2002 pushed a commit that referenced this pull request Dec 19, 2025
yueming-yuan pushed a commit to yueming-yuan/slime that referenced this pull request Dec 29, 2025
Yangruipis pushed a commit to redai-infra/slime that referenced this pull request Feb 28, 2026
mouad-hpc pushed a commit to mouad-hpc/slime that referenced this pull request Mar 24, 2026
Revert "fix: run CI steps directly in slimerl/slime pod with native GPU access"
jamescsq47 pushed a commit to jamescsq47/slime that referenced this pull request Mar 28, 2026
…-black

Format convert_fsdp_to_hf with black
HJSang added a commit to HJSang/slime that referenced this pull request Jun 14, 2026
- #1 add --opsd-offload-teacher-logits to offload full-vocab teacher logits to CPU
  between forwards (moved back to device per micro-batch in the loss); keep the
  on-GPU view by default. Chunked JSD remains future work (THUDM#4, deferred).
- THUDM#2 warn when privileged_info is empty/None (teacher==student context -> ~0 signal).
- THUDM#3 repack teacher micro-batches by teacher lengths on the actor side, keeping the
  student's sample-to-rank assignment so response positions stay aligned
  (repack_micro_batches_by_length in dp_schedule.py; forward-only, per-rank).
- THUDM#5 validate OPSD is not combined with --disable-compute-advantages-and-returns.
- THUDM#6 skip loading the ref model under OPSD (never forwarded).
- THUDM#7 descriptive error when a dataset row lacks the privileged-info field.

Tests: add repack unit tests (coverage, token budget, oversized-sample-alone).
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants