Merge pull request #1249 from rohithreddy0087/patch-1

Remove draft_neuron_config.sequence_parallel_enabled flag in 405b eagle speculation example to avoid compilation error

Author: chongmni-aws

libraries/nxd-inference/tutorials/trn2-llama3.1-405b-tutorial.rst

@@ -303,7 +303,6 @@ This example uses the following configuration options:
        draft_neuron_config.trace_tokengen_model = True
        draft_neuron_config.enable_fused_speculation = False
        draft_neuron_config.is_eagle_draft = True
-       draft_neuron_config.sequence_parallel_enabled = False
        draft_config = LlamaInferenceConfig(
            draft_neuron_config,
            load_config=load_pretrained_config(draft_model_path)
@@ -358,4 +357,4 @@ This example uses the following configuration options:
    
    
    if __name__ == "__main__":
-       run_llama_generate()
+       run_llama_generate()