update README.md

README.md

@@ -24,11 +24,12 @@ tags:
 ## 🔥 Why Need PixArt-LCM
 Following [LCM LoRA](https://huggingface.co/blog/lcm_lora), we illustrative of the generation speed we achieve on various computers. Let us stress again how liberating it is to explore image generation so easily with PixArt-LCM.
 
-| Hardware                    | PixArt-LCM (4 steps)   | SDXL LoRA LCM (4 steps) | PixArt standard (14 steps) | SDXL standard (25 steps) |
-|-----------------------------|------------------------|-------------------------|----------------------------|---------------------------|
-| T4 (Google Colab Free Tier) | 3.3s                   | 8.4s                    | 16.0s                      | 26.5s                     |
-| A100 (80 GB)                | 0.51s                  | 1.2s                    | 2.2s                       | 3.8s                      |
-| V100 (32 GB)                | 1.2s                   | 1.2s                    | 5.5s                       | 7.7s                      |
+| Hardware                    | PixArt-LCM (4 steps) | SDXL LoRA LCM (4 steps) | PixArt standard (14 steps) | SDXL standard (25 steps) |
+|-----------------------------|----------------------|-------------------------|----------------------------|---------------------------|
+| T4 (Google Colab Free Tier) | 3.3s                 | 8.4s                    | 16.0s                      | 26.5s                     |
+| A100 (80 GB)                | 0.51s                | 1.2s                    | 2.2s                       | 3.8s                      |
+| V100 (32 GB)                | 0.8s                 | 1.2s                    | 5.5s                       | 7.7s                      |
+
 These tests were run with a batch size of 1 in all cases.
 
 For cards with a lot of capacity, such as A100, performance increases significantly when generating multiple images at once, which is usually the case for production workloads.