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 c7ffadf commit cfcb97aCopy full SHA for cfcb97a
tensorrt_llm/llmapi/llm_args.py
@@ -1433,7 +1433,7 @@ def validate_speculative_config(self):
1433
self.build_config.max_draft_len = self.speculative_config.max_draft_len
1434
self.build_config.speculative_decoding_mode = SpeculativeDecodingMode.EAGLE
1435
if self.speculative_config.eagle3_one_model:
1436
- self.num_extra_kv_tokens = self.max_draft_len - 1
+ self.speculative_config.num_extra_kv_tokens = self.speculative_config.max_draft_len - 1
1437
if self.backend not in ['pytorch', '_autodeploy']:
1438
eagle_config = _EagleConfig(
1439
self.speculative_config.eagle_choices,
0 commit comments