Skip to content

Fix gqa model tflops compute #1099

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

Merged
zhuzilin merged 1 commit into from
Dec 12, 2025
Merged

Fix gqa model tflops compute #1099

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: 8 additions & 4 deletions slime/utils/flops_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,8 +16,8 @@ def calculate_qkv_projection_flops(seqlen, hidden_size, num_attention_heads, num


def calculate_attention_flops(seqlen, num_attention_heads, head_dim):
# QK^T
flops = 2 * num_attention_heads * seqlen * seqlen * head_dim
# QK^T with causal
flops = 2 * num_attention_heads * seqlen * seqlen * head_dim // 2
# A*V
flops += 2 * num_attention_heads * seqlen * seqlen * head_dim
return flops
Expand All @@ -31,8 +31,9 @@ def calculate_mlp_flops(seqlen, hidden_size, ffn_hidden_size):
return 2 * seqlen * hidden_size * ffn_hidden_size * 3


def calculate_layer_flops(seqlen, hidden_size, num_attention_heads, num_query_groups, ffn_hidden_size):
head_dim = hidden_size // num_attention_heads
def calculate_layer_flops(seqlen, hidden_size, num_attention_heads, num_query_groups, ffn_hidden_size, head_dim):
if head_dim is None:
head_dim = hidden_size // num_attention_heads
return (
calculate_qkv_projection_flops(seqlen, hidden_size, num_attention_heads, num_query_groups)
+ calculate_attention_flops(seqlen, num_attention_heads, head_dim)
Expand All @@ -49,6 +50,7 @@ def calculate_fwd_flops(
num_attention_heads = args.num_attention_heads
num_query_groups = args.num_query_groups
vocab_size = args.vocab_size
kv_channels = args.kv_channels

total_flops = 0

Expand Down Expand Up @@ -82,6 +84,7 @@ def calculate_fwd_flops(
num_attention_heads,
num_query_groups,
dense_ffn,
kv_channels,
)
* num_dense_layers
)
Expand All @@ -94,6 +97,7 @@ def calculate_fwd_flops(
num_attention_heads,
num_query_groups,
moe_ffn,
kv_channels,
)
* num_moe_layers
)
Expand Down