Update README.md

README.md

@@ -50,14 +50,14 @@ Alternatively, you can use NeMo Megatron training docker container with all depe
 
 ### Step 2: Launch eval server 
 
-**Note.** The example below launches a model variant with Tensor Parallelism (TP) of 8 and Pipeline Parallelism (PP) of 1 on 8 GPUs.
+**Note.** The example below launches a model variant with Tensor Parallelism (TP) of 4 and Pipeline Parallelism (PP) of 1 on 4 GPUs.
 
 
 ```
 git clone https://github.com/NVIDIA/NeMo.git 
 cd NeMo/examples/nlp/language_modeling
 git checkout v1.11.0
-python megatron_gpt_eval.py gpt_model_file=nemo_gpt20B_bf16_tp8.nemo server=True tensor_model_parallel_size=2 trainer.devices=2
+python megatron_gpt_eval.py gpt_model_file=nemo_gpt20B_bf16_tp4.nemo server=True tensor_model_parallel_size=4 trainer.devices=4
 ```
 
 ### Step 3: Send prompts to you model!