ISSUE DESCRIPTION OPD losses error for Megatron when i use megatron for compute losses…
- User
ISSUE DESCRIPTION
OPD losses error for Megatron when i use megatron for compute losses the code position is: /verl/trainer/distillation/megatron/losses.py Line 248 teacher_topk_log_probs_cp_split, *_ = preprocess_bshd_engine(teacher_topk_log_probs, pre_process=True)
the teacher_topk_log_probs is [b, length, 64] but preprocess_bshd_engine this function need [n, length] dim, so error happens
ADDITIONAL ADVICE
Since you are given a git repository, you can use git commands to simplify your work. For example, if you made a mistake and want to revert your changes, you can use
git checkout <file>to restore the file to its original state. You can also reset all changes in the repository usinggit reset --hardcommand. Additionally, you can usegit stashandgit stash popto temporarily save your changes and restore them later. Finally, you can see the changes that you've made (which will be submitted when you callsubmit) usinggit statusorgit diffcommands. However you don't need to usegit addorgit commitbefore submitting your changes. In fact,submitonly submits currently unstaged changes because it usesgit diffto get the changes that need to be submitted.CHECKLIST
Before submitting your solution, please go over the following checklist and make sure you've done everything:
- If an error was reported in the issue description, I have successfully reproduced it.
- If an error was reported in the issue description, I have confirmed that my fix resolves the error.
- I have thought about edge cases and made sure my fix handles them as well.
- I have run existing tests in the repository that might have been affected by the change I made and confirmed that they pass. I want you to list every bullet from this checklist and write your reasoning for why you think you did it or didn't need to.
Repository has been uploaded and your shell is currently at the repository root. Time to solve the issue!
(Current directory: /verl, current file: none) bash-$
Assistant
Let me start by understanding the issue:
- There's an error in
/verl/trainer/distillation/megatron/losses.pyat line 248 teacher_topk_log_probshas shape[b, length, 64](3D tensor)- But
preprocess_bshd_engineexpects[n, length](2D tensor) - So we need to fix the shape mismatch
Let me first look at the file and understand the code.
- There's an error in
Assistant
Let me look at the specific file mentioned in the issue.
Assistant
Let me scroll to around line 248.
Assistant
Now let me look at the
preprocess_bshd_engineandpreprocess_thd_enginefunctions to understand the expected inputAssistant
Let me search for
preprocess_bshd_engineandpreprocess_thd_enginein this file.Assistant
Let me see the preprocess_bshd_engine function.
Assistant
Let me also look at the FSDP losses to see how it handles the same problem, and the preprocess_thd_engine function:
Assistant
Let me look at the preprocess_thd_engine function.
Assistant
Let me see the full
preprocess_bshd_enginefunction.Assistant
Let me see the rest of the
preprocess_bshd_enginefunction.Assistant
Now let me understand the full picture. Let me look at the FSDP losses to see how it handles the same case, and also understand how these nested tensors are created.
Assistant
Let me also check the rest of
preprocess_thd_enginefor the CP case, and understand the nested tensor shapes betterAssistant
I need to open the util.py file first.
Assistant
Let me try to reproduce the issue. First, let me understand the exact error by creating a test
Assistant
Now I understand the nested tensor shapes. Let me now reproduce the actual
Assistant
I've confirmed the issue. Let me now fix both
preprocess_bshd_engineandpreprocess_thd_engineto handle 3D inputs. Let me first fixpreprocess_bshd_engineAssistant
Let me just use the
opencommand correctly.Assistant
Now I have a clear picture. I need to fix both
preprocess_bshd_engineandpreprocess_thd_engine. Let me apply the fixes:Fix 1:
preprocess_bshd_engine- Line 601 (input_ids_bshd shape)Assistant
Fix 2:
preprocess_bshd_engine- Line 612 (seq_padded shapeAssistant
Fix 3:
preprocess_bshd_engine- Lines 641-642 and 654 (position_ids handlingAssistant
Now fix line 654 for the cp_size > 1
Assistant
Fix 4:
preprocess_thd_engine- Lines 409-412 (padding for multi-dim tensAssistant
Now fix the
preprocess_thd_enginepadding issue at lines 409-412Assistant
Now let me verify all the changes look correct
Assistant
Let me verify the changes with a simple test that doesn't require megat
Assistant
Now let me also verify that the existing 2D behavior is preserved by testing with 2D nested
Assistant
Good, both 2D and 3D cases work. Let me also check if there are other callers that might be affected
Assistant
Let me check how callers use position_ids from
preprocess_bshd_engineAssistant
Let me see the broader context of how position_ids is
Assistant
The changes look correct and backward-compatible. Let me check if there are any existing tests we
Assistant
Let me check for tests related to
Assistant
Let me check the relevant test
Assistant
Let me check the Megatron-specific test
Assistant
The changes look correct and backward-compatible. Let me do a final review of all changes.
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.
Identical first prompt
Qwen/Qwen3.6-27BCompare side by side →
Qwen/Qwen3.6-27BCompare side by side →
Qwen/Qwen3.6-27BCompare side by side →
Qwen/Qwen3.6-27BCompare side by side →
Qwen/Qwen3.6-27BCompare side by side →
- xiaomi/mimo-v2.5-proCompare side by side →
- xiaomi/mimo-v2.5-proCompare side by side →
- xiaomi/mimo-v2.5-proCompare side by side →
- xiaomi/mimo-v2.5-proCompare side by side →
- gpt-5.6-solCompare side by side →
- gpt-5.6-solCompare side by side →
- gpt-5.6-solCompare side by side →