huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-06-19 15:52:08 +00:00
Code Issues Packages Projects Releases Wiki Activity

Remove debug info

Browse Source 
Signed-off-by: yuanwu <yuan.wu@intel.com>
This commit is contained in:
yuanwu 2025-06-06 06:17:45 +00:00
parent 7f346a88e3
commit eed58b77c3
2 changed files with 0 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
backends/gaudi/server/text_generation_server/models/custom_modeling/flash_qwen3_moe_modeling.py
View File

@ -229,21 +229,8 @@ class Qwen3MoE(nn.Module):
self.process_group = weights.process_group
def forward(self, x: torch.Tensor) -> torch.Tensor:
# router_logits: (num_tokens, n_experts)
router_logits = self.gate(x)
# synchronize(x.device)
# real_free_memory = get_free_memory(x.device, 1)
# log_master(
# logger.debug,
# f"moe forward 1Free memory real: {real_free_memory / 1e9:.2f}GB"
# )
out = self.moe(x, gating_output=router_logits)
# synchronize(x.device)
# real_free_memory = get_free_memory(x.device, 1)
# log_master(
# logger.debug,
# f"moe forward 2 Free memory real: {real_free_memory / 1e9:.2f}GB"
# )
# Reduce sum
if self.process_group.size() > 1:

1
backends/gaudi/server/text_generation_server/models/flash_causal_lm.py
View File

@ -1412,7 +1412,6 @@ class FlashCausalLM(Model):
aliases=aliases,
weights_loader=weights_loader,
)
print(f"weights: {weights}")
prefix = None
model = model_class(prefix, config, weights)