[float8 moe training] validate float8 moe parallelism config #1360
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
danielvegamyhre
requested review from
tianyu-l,
fegin,
wwwjn and
wconstab
as code owners
facebook-github-bot
added
the
CLA Signed
mori360
pushed a commit
to mori360/torchtitan
that referenced
this pull request
Jul 8, 2025
…#1360) ## Summary Validate only FSDP, HSDP are used for float8 MoE training. TP support is in progress and CP/PP are untested. 2D+ parallelism are untested as well. ## Test plan - Command: `NGPU=4 CONFIG_FILE="./torchtitan/experiments/llama4/train_configs/debug_model.toml" ./run_train.sh --training.steps=10 --model.converters="float8" --float8.recipe_name="rowwise" --float8.moe_fqns_prototype="experts" --parallelism.tensor_parallel_degree=2` - Error: `AssertionError: Float8 MoE training prototype does not yet support tensor parallelism`
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Outdated suggestions cannot be applied.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.
Summary
Validate only FSDP, HSDP are used for float8 MoE training. TP support is in progress and CP/PP are untested. 2D+ parallelism are untested as well.
Test plan
NGPU=4 CONFIG_FILE="./torchtitan/experiments/llama4/train_configs/debug_model.toml" ./run_train.sh --training.steps=10 --model.converters="float8" --float8.recipe_name="rowwise" --float8.moe_fqns_prototype="experts" --parallelism.tensor_parallel_degree=2AssertionError: Float8 MoE training prototype does not yet support tensor parallelism