Skip to content

[Feature] Implement multi MTP Loss #2

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
x54-729 wants to merge 9 commits into
base: mtp
Choose a base branch
from
Open

[Feature] Implement multi MTP Loss #2

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
abc64dd
[Refactor] Refactor loss context API to support multiple loss types
HAOCHENYE Mar 24, 2026
18e64ae
[Feature] Add Multi-Token Prediction (MTP) module implementation
HAOCHENYE Mar 24, 2026
7a9d596
[Refactor] Rename CELossContext to LMHeadLossContext and refactor los…
HAOCHENYE Mar 24, 2026
ed9b0b3
[Feature] Add SP support for roll_sequence_context in MTP
HAOCHENYE Mar 24, 2026
6284035
implement multi mtp loss for sci sft
x54-729 Mar 25, 2026
f641d6e
fix typos
x54-729 Mar 25, 2026
d4904a8
fix mtp_forward and mtp_ctx after rebase
x54-729 Mar 25, 2026
283ba18
small fix
x54-729 Mar 25, 2026
c856fd9
Fix MLLMPretrainHybridPackDataset (#1638)
HAOCHENYE Mar 30, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions tests/engine/test_dense_train_engine.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -83,13 +83,13 @@ def warmup_fn(x):
seq_ctx = seq_ctx.split(sequence_parallel_mesh=sp_mesh)
seq_ctx_list = [seq_ctx]
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=labels, sp_mesh=sp_mesh)
loss_ctx = loss_cfg.build(data={"shifted_labels": labels}, sp_mesh=sp_mesh)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)

seq_ctx = seq_ctx_list[0]
loss_ctx = loss_ctx_list[0]
engine_input = [ModelItem(seq_ctx=seq_ctx, loss_ctx=loss_ctx)]
engine_input = [ModelItem(seq_ctx=seq_ctx, loss_ctx={"lm": loss_ctx})]
loss_log = engine.train_step(engine_input)["logs_info"]
grad_norm = engine.clip_grad_norm()
engine.step_optimizer(grad_norm)
Expand Down
8 changes: 4 additions & 4 deletions tests/engine/test_moe_train_engine.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -93,12 +93,12 @@ def warmup_fn(x):
seq_ctx.num_padding = pack_len
seq_ctx_list = [seq_ctx]
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=labels, sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": labels}, sp_mesh=None)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
seq_ctx = seq_ctx_list[0]
engine_input = [ModelItem(seq_ctx=seq_ctx, loss_ctx=loss_ctx)]
engine_input = [ModelItem(seq_ctx=seq_ctx, loss_ctx={"lm": loss_ctx})]
loss_log = engine.train_step(engine_input)["logs_info"]
grad_norm = engine.clip_grad_norm()
engine.step_optimizer(grad_norm)
Expand Down Expand Up @@ -184,12 +184,12 @@ def warmup_fn(x):
seq_ctx.num_padding = pack_len
seq_ctx_list = [seq_ctx]
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=labels, sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": labels}, sp_mesh=None)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
seq_ctx = seq_ctx_list[0]
engine_input = [ModelItem(seq_ctx=seq_ctx, loss_ctx=loss_ctx)]
engine_input = [ModelItem(seq_ctx=seq_ctx, loss_ctx={"lm": loss_ctx})]
loss_log = engine.train_step(engine_input)["logs_info"]
grad_norm = engine.clip_grad_norm()
engine.step_optimizer(grad_norm)
Expand Down
12 changes: 6 additions & 6 deletions tests/engine/test_moe_train_engine_float8.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -87,12 +87,12 @@ def warmup_fn(x):
seq_ctx.num_padding = pack_len
seq_ctx_list = [seq_ctx]
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=labels, sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": labels}, sp_mesh=None)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
seq_ctx = seq_ctx_list[0]
engine_input = [ModelItem(seq_ctx=seq_ctx, loss_ctx=loss_ctx)]
engine_input = [ModelItem(seq_ctx=seq_ctx, loss_ctx={"lm": loss_ctx})]
loss_log = engine.train_step(engine_input)["logs_info"]
grad_norm = engine.clip_grad_norm()
engine.step_optimizer(grad_norm)
Expand Down Expand Up @@ -165,12 +165,12 @@ def warmup_fn(x):
seq_ctx.num_padding = pack_len
seq_ctx_list = [seq_ctx]
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=labels, sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": labels}, sp_mesh=None)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
seq_ctx = seq_ctx_list[0]
engine_input = [ModelItem(seq_ctx=seq_ctx, loss_ctx=loss_ctx)]
engine_input = [ModelItem(seq_ctx=seq_ctx, loss_ctx={"lm": loss_ctx})]
loss_log = engine.train_step(engine_input)["logs_info"]
grad_norm = engine.clip_grad_norm()
engine.step_optimizer(grad_norm)
Expand Down Expand Up @@ -264,12 +264,12 @@ def warmup_fn(x):
seq_ctx.to('cuda')
seq_ctx_list = [seq_ctx]
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=labels, sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": labels}, sp_mesh=None)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
seq_ctx = seq_ctx_list[0]
engine_input = [ModelItem(seq_ctx=seq_ctx, loss_ctx=loss_ctx)]
engine_input = [ModelItem(seq_ctx=seq_ctx, loss_ctx={"lm": loss_ctx})]
logs_info = engine.train_step(engine_input)["logs_info"]
grad_norm = engine.clip_grad_norm()
engine.step_optimizer(grad_norm)
Expand Down
8 changes: 4 additions & 4 deletions tests/loss/test_ce_loss.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,7 +72,7 @@ def test_global_loss_reduction(self, loss_mode, grad_accumulation_steps, chunk_s
for data in data_batch:
seq_ctx = data["seq_ctx"]
seq_ctx_list.append(seq_ctx)
loss_ctx = loss_cfg.build(shifted_labels=data["shifted_labels"], sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": data["shifted_labels"]}, sp_mesh=None)
loss_ctx_list.append(loss_ctx)
loss_ctx_list = CELossContext.build_batches(loss_ctx_list, cu_seq_lens_list=[seq_ctx.cu_seq_lens_q for seq_ctx in seq_ctx_list])

Expand Down Expand Up @@ -172,7 +172,7 @@ def test_other_loss_reduction(self, loss_reduction, loss_mode, grad_accumulation
for data in data_batch:
seq_ctx = data["seq_ctx"]
seq_ctx_list.append(seq_ctx)
loss_ctx = loss_cfg.build(shifted_labels=data["shifted_labels"], sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": data["shifted_labels"]}, sp_mesh=None)
loss_ctx_list.append(loss_ctx)
loss_ctx_list = CELossContext.build_batches(loss_ctx_list, cu_seq_lens_list=[seq_ctx.cu_seq_lens_q for seq_ctx in seq_ctx_list])

Expand Down Expand Up @@ -310,7 +310,7 @@ def test_sp_global_loss_reduction(self, loss_mode, sp_size, grad_accumulation_st
sp_mesh = data_mesh['sp']
seq_ctx.sequence_parallel_mesh = sp_mesh
seq_ctx_list = [seq_ctx]
loss_ctx = loss_cfg.build(shifted_labels=target, sp_mesh=sp_mesh)
loss_ctx = loss_cfg.build(data={"shifted_labels": target}, sp_mesh=sp_mesh)
loss_ctx_list = [loss_ctx]
if sp_size > 1:
seq_ctx_list[0] = seq_ctx_list[0].split(sequence_parallel_mesh=sp_mesh)
Expand Down Expand Up @@ -397,7 +397,7 @@ def test_sp_others_loss_reduction(self, loss_reduction, loss_mode, sp_size, grad
sp_mesh = data_mesh['sp']
seq_ctx.sequence_parallel_mesh = sp_mesh
seq_ctx_list = [seq_ctx]
loss_ctx = loss_cfg.build(shifted_labels=target, sp_mesh=sp_mesh)
loss_ctx = loss_cfg.build(data={"shifted_labels": target}, sp_mesh=sp_mesh)
loss_ctx_list = [loss_ctx]
if sp_size > 1:
seq_ctx_list[0] = seq_ctx_list[0].split(sequence_parallel_mesh=sp_mesh)
Expand Down
2 changes: 1 addition & 1 deletion tests/loss/test_grpo_loss.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -147,7 +147,7 @@ def test_grpo_loss(self, grad_acc, sp_size, kl_loss_coef, loss_mode, chunk_size,
if sp_size > 1:
seq_ctx = seq_ctx.split(sp_mesh)
seq_ctx_list.append(seq_ctx)
loss_ctx = loss_cfg.build(shifted_labels=shifted_labels_list_rank[iter_idx], advantages=advantages_list_rank[iter_idx], sp_mesh=sp_mesh)
loss_ctx = loss_cfg.build(data={"shifted_labels": shifted_labels_list_rank[iter_idx], "advantages": advantages_list_rank[iter_idx]}, sp_mesh=sp_mesh)
loss_ctx_list.append(loss_ctx)

with torch.no_grad():
Expand Down
3 changes: 1 addition & 2 deletions tests/loss/test_oreal_loss.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -216,8 +216,7 @@ def test_grpo_loss(self, grad_acc, sp_size, kl_loss_coef, loss_mode, chunk_size,
seq_ctx = seq_ctx.split(sp_mesh)
seq_ctx_list.append(seq_ctx)
loss_ctx = loss_cfg.build(
shifted_labels=shifted_labels_list_rank[iter_idx],
advantages=advantages_list_rank[iter_idx],
data={"shifted_labels": shifted_labels_list_rank[iter_idx], "advantages": advantages_list_rank[iter_idx]},
sp_mesh=sp_mesh,
)
loss_ctx_list.append(loss_ctx)
Expand Down
8 changes: 4 additions & 4 deletions tests/model/test_gpt_oss_moe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -78,7 +78,7 @@ def test_gpt_oss_run(self, device, dispatcher, ep_size, compile, tol, loss_class
loss_cfg = CELossConfig()
seq_ctx_list = [seq_ctx]
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=shifted_labels, sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": shifted_labels}, sp_mesh=None)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
Expand All @@ -87,7 +87,7 @@ def test_gpt_oss_run(self, device, dispatcher, ep_size, compile, tol, loss_class
with torch.no_grad():
output = gpt_oss_model(
seq_ctx=seq_ctx,
loss_ctx=loss_ctx,
loss_ctx={"lm": loss_ctx},
)
loss = output["loss"]
self.assertTrue(torch.allclose(loss, expected_loss.to(loss.dtype), atol=tol, rtol=tol))
Expand Down Expand Up @@ -141,7 +141,7 @@ def test_fsdp_accuracy(self, device, dispatcher, ep_size):
loss_cfg = CELossConfig()
seq_ctx_list = [seq_ctx]
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=shifted_labels, sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": shifted_labels}, sp_mesh=None)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
Expand All @@ -152,7 +152,7 @@ def test_fsdp_accuracy(self, device, dispatcher, ep_size):
with torch.no_grad():
output = gpt_oss_model(
seq_ctx=seq_ctx,
loss_ctx=loss_ctx,
loss_ctx={"lm": loss_ctx},
)
loss = output["loss"]
self.assertTrue(torch.allclose(loss, expected_loss.to(loss.dtype), atol=1e-2, rtol=1e-2))
Expand Down
16 changes: 8 additions & 8 deletions tests/model/test_intern_s1.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -78,7 +78,7 @@ def test_interns1_text_run(self, device, tol):

seq_ctx_list = [seq_ctx]
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=shifted_labels, sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": shifted_labels}, sp_mesh=None)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
Expand All @@ -87,7 +87,7 @@ def test_interns1_text_run(self, device, tol):
with torch.no_grad():
output = interns1_model(
seq_ctx=seq_ctx,
loss_ctx=loss_ctx,
loss_ctx={"lm": loss_ctx},
)
loss = output["loss"]
self.assertTrue(torch.allclose(loss, expected_loss.to(loss.dtype), atol=tol, rtol=tol))
Expand Down Expand Up @@ -186,7 +186,7 @@ def test_interns1_image_run(self, device, sp_size, tol):

seq_ctx_list = [seq_ctx]
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=shifted_labels, sp_mesh=sp_mesh)
loss_ctx = loss_cfg.build(data={"shifted_labels": shifted_labels}, sp_mesh=sp_mesh)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
Expand All @@ -195,7 +195,7 @@ def test_interns1_image_run(self, device, sp_size, tol):
with torch.no_grad():
output = interns1_model(
seq_ctx=seq_ctx,
loss_ctx=loss_ctx,
loss_ctx={"lm": loss_ctx},
)
loss = output["loss"]
self.assertTrue(torch.allclose(loss, expected_loss.to(loss.dtype), atol=tol, rtol=tol))
Expand Down Expand Up @@ -256,7 +256,7 @@ def test_fsdp_text_accuracy(self, device, tol):
seq_ctx_list = [seq_ctx]
loss_cfg = CELossConfig()
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=shifted_labels, sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": shifted_labels}, sp_mesh=None)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
Expand All @@ -265,7 +265,7 @@ def test_fsdp_text_accuracy(self, device, tol):
with torch.no_grad():
output = interns1_model(
seq_ctx=seq_ctx,
loss_ctx=loss_ctx,
loss_ctx={"lm": loss_ctx},
)
loss = output["loss"]
self.assertTrue(torch.allclose(loss, expected_loss.to(loss.dtype), atol=tol, rtol=tol))
Expand Down Expand Up @@ -370,7 +370,7 @@ def test_fsdp_image_accuracy(self, device, sp_size, compile, tol):
seq_ctx_list = [seq_ctx]
loss_cfg = CELossConfig()
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=shifted_labels, sp_mesh=sp_mesh)
loss_ctx = loss_cfg.build(data={"shifted_labels": shifted_labels}, sp_mesh=sp_mesh)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
Expand All @@ -379,7 +379,7 @@ def test_fsdp_image_accuracy(self, device, sp_size, compile, tol):
with torch.no_grad():
output = interns1_model(
seq_ctx=seq_ctx,
loss_ctx=loss_ctx,
loss_ctx={"lm": loss_ctx},
)
loss = output["loss"]
self.assertTrue(torch.allclose(loss, expected_loss.to(loss.dtype), atol=tol, rtol=tol))
Expand Down
10 changes: 5 additions & 5 deletions tests/model/test_moe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -62,12 +62,12 @@ def test_moe_config(self, dtype, device):

seq_ctx_list = [seq_ctx]
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=shifted_labels, sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": shifted_labels}, sp_mesh=None)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
seq_ctx = seq_ctx_list[0]
model(seq_ctx=seq_ctx, loss_ctx=loss_ctx)
model(seq_ctx=seq_ctx, loss_ctx={"lm": loss_ctx})


class TestDistributedMoE(DeterministicDDPTestCase):
Expand Down Expand Up @@ -135,15 +135,15 @@ def test_parallel_accuracy(self, dtype, device, dispatcher, n_shared_experts, fi

seq_ctx_list = [seq_ctx]
LossContext = loss_cfg.loss_ctx_cls
loss_ctx = loss_cfg.build(shifted_labels=shifted_labels, sp_mesh=None)
loss_ctx = loss_cfg.build(data={"shifted_labels": shifted_labels}, sp_mesh=None)
loss_ctx_list = [loss_ctx]
loss_ctx_list = LossContext.build_batches(loss_ctx_list)
loss_ctx = loss_ctx_list[0]
seq_ctx = seq_ctx_list[0]

loss_parallel = parallel_model(seq_ctx=seq_ctx, loss_ctx=loss_ctx)["loss"]
loss_parallel = parallel_model(seq_ctx=seq_ctx, loss_ctx={"lm": loss_ctx})["loss"]

loss_expected = model(seq_ctx=seq_ctx, loss_ctx=loss_ctx)["loss"]
loss_expected = model(seq_ctx=seq_ctx, loss_ctx={"lm": loss_ctx})["loss"]

torch.allclose(loss_expected, loss_parallel, atol=1e-6, rtol=1e-4)

Expand Down
Loading