move PP API to model agnostic file #1868
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
meta-cla
bot
added
the
CLA Signed
tianyu-l
force-pushed
the
pipeline_refactor
branch
from
87b088d to
8f70bbf
Compare
githubsgi
pushed a commit
to githubsgi/torchtitan
that referenced
this pull request
Oct 15, 2025
We originally thought each model should have its own `pipeline.py` function. However, for most LLMs, it turns out a single function would suffice, and all models which needs PP are reusing `pipeline_llama.py` originally written for llama3. (For diffusion models, the model size doesn't justify the usage of PP.) This PR consolidates them and moves `pipeline_llm` into `torchtitan/distributed/pipeline_parallel.py`. We can refactor later if things change.
githubsgi
pushed a commit
to githubsgi/torchtitan
that referenced
this pull request
Oct 16, 2025
We originally thought each model should have its own `pipeline.py` function. However, for most LLMs, it turns out a single function would suffice, and all models which needs PP are reusing `pipeline_llama.py` originally written for llama3. (For diffusion models, the model size doesn't justify the usage of PP.) This PR consolidates them and moves `pipeline_llm` into `torchtitan/distributed/pipeline_parallel.py`. We can refactor later if things change.
githubsgi
pushed a commit
to githubsgi/torchtitan
that referenced
this pull request
Oct 29, 2025
We originally thought each model should have its own `pipeline.py` function. However, for most LLMs, it turns out a single function would suffice, and all models which needs PP are reusing `pipeline_llama.py` originally written for llama3. (For diffusion models, the model size doesn't justify the usage of PP.) This PR consolidates them and moves `pipeline_llm` into `torchtitan/distributed/pipeline_parallel.py`. We can refactor later if things change.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
We originally thought each model should have its own
pipeline.pyfunction.However, for most LLMs, it turns out a single function would suffice, and all models which needs PP are reusing
pipeline_llama.pyoriginally written for llama3.(For diffusion models, the model size doesn't justify the usage of PP.)
This PR consolidates them and moves
pipeline_llmintotorchtitan/distributed/pipeline_parallel.py.We can refactor later if things change.