Update README.md

README.md

@@ -133,7 +133,7 @@ llamafactory-cli train logicsct_train_Phi4_qlora_sft_otfq.yaml       # VRAM used
 llamafactory-cli chat logicsct_inference_Phi4_qlora_sft_otfq.yaml    # VRAM used: 30927MiB for inference of base model + QLoRA adapter
 llamafactory-cli export logicsct_export_Phi4_qlora_sft.yaml          # VRAM used:   665MiB + about 29 GB of system RAM for exporting a merged verison of the model with its adapter
 llamafactory-cli export logicsct_export_Phi4_qlora_sft_Q4.yaml       # VRAM used: 38277MiB for a 4bit quant export of the merged model
-llamafactory-cli chat logicsct_inference_Phi4_qlora_sft_otfq_Q4.yaml # VRAM used:  9255MiB-11405MiB VRAM for inference of the 4bit quant merged model (increasing with increasing context length)
+llamafactory-cli chat logicsct_inference_Phi4_qlora_sft_otfq_Q4.yaml # VRAM used:  9255MiB-11405MiB for inference of the 4bit quant merged model (increasing with increasing context length)
 ```
 
 ### Comparison of Open Source Training/Models with OpenAI Proprietary Fine-Tuning