Skip to content

Fix dtype leaks that break bfloat16 training #26

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
tonyzdev wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Fix dtype leaks that break bfloat16 training #26

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
12 changes: 6 additions & 6 deletions open_mythos/main.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -823,7 +823,7 @@ def forward(
B, T, D = h.shape

halted = torch.zeros(B, T, device=h.device, dtype=torch.bool)
cumulative_p = torch.zeros(B, T, device=h.device)
cumulative_p = torch.zeros(B, T, device=h.device, dtype=h.dtype)
h_out = torch.zeros_like(h)

for t in range(n_loops):
Expand All @@ -848,10 +848,10 @@ def forward(
remainder,
p,
)
weight = weight * still_running.float()
weight = weight * still_running.to(h.dtype)
h_out = h_out + weight.unsqueeze(-1) * h

cumulative_p = cumulative_p + p * still_running.float()
cumulative_p = cumulative_p + p * still_running.to(h.dtype)
halted = halted | (cumulative_p >= self.cfg.act_threshold)

# Only short-circuit when there is no KV cache to keep consistent.
Expand Down Expand Up @@ -938,7 +938,7 @@ def _init_weights(self) -> None:
nn.init.normal_(m.weight, std=0.02)

@staticmethod
def _causal_mask(seq_len: int, device: torch.device) -> torch.Tensor:
def _causal_mask(seq_len: int, device: torch.device, dtype: torch.dtype = torch.float32) -> torch.Tensor:
"""
Build an additive causal mask: 0 on and below the diagonal, -inf above.

Expand All @@ -949,7 +949,7 @@ def _causal_mask(seq_len: int, device: torch.device) -> torch.Tensor:
Returns:
Tensor of shape (1, 1, seq_len, seq_len) broadcastable over (B, H, T, S)
"""
mask = torch.full((1, 1, seq_len, seq_len), float("-inf"), device=device)
mask = torch.full((1, 1, seq_len, seq_len), float("-inf"), device=device, dtype=dtype)
return torch.triu(mask, diagonal=1)

def forward(
Expand Down Expand Up @@ -983,7 +983,7 @@ def forward(
freqs_cis = (
self.freqs_cis_mla if self.cfg.attn_type == "mla" else self.freqs_cis
)[start_pos : start_pos + T]
mask = self._causal_mask(T, device) if T > 1 else None
mask = self._causal_mask(T, device, dtype=x.dtype) if T > 1 else None

for i, layer in enumerate(self.prelude):
x = layer(x, freqs_cis, mask, kv_cache, cache_key=f"prelude_{i}")
Expand Down