We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent 6a638aa commit 658d463Copy full SHA for 658d463
tensorrt_llm/_torch/model_config.py
@@ -426,8 +426,10 @@ def from_pretrained(cls,
426
trust_remote_code=trust_remote_code,
427
**kwargs,
428
)
429
- if pretrained_config.architectures[0] == "DeepseekV32ForCausalLM":
430
- sparse_attention_config = kwargs.get('sparse_attention_config')
+ if pretrained_config.architectures[
+ 0] == "DeepseekV32ForCausalLM":
431
+ sparse_attention_config = kwargs.get(
432
+ 'sparse_attention_config')
433
if sparse_attention_config:
434
index_n_heads = sparse_attention_config.index_n_heads or pretrained_config.index_n_heads
435
index_head_dim = sparse_attention_config.index_head_dim or pretrained_config.index_head_dim
0 commit comments