Update README.md (#88)

Author: nicholasachow

README.md

@@ -71,11 +71,11 @@ See [example](flexgen/apps/data_wrangle)
 The corresponding effective batch sizes are in parentheses. Please see [here](benchmark/batch_size_table.md) for more details.
 | System | OPT-6.7B | OPT-30B | OPT-175B |
 | ------ | -------- | ------- | -------- |
-| Hugging Face Accelerate   | 25.12 (2 on gpu) | 0.62 (8 on cpu	) | 0.01 (2 on disk) |
-| DeepSpeed ZeRO-Inference | 9.28 (16 on cpu)  | 0.60 (4 on cpu) | 0.01 (1 on disk) |
-| Petals                 | 8.25     | 2.84    | 0.08 |
-| FlexGen                  | 25.26 (2 on gpu) | 7.32 (144 on cpu) | 0.69 (256 on disk) |
-| FlexGen with Compression | **29.12** (72 on gpu) | **8.38** (512 on cpu) | **1.12** (144 on cpu) |
+| Hugging Face Accelerate   | 25.12 (2 on GPU) | 0.62 (8 on CPU) | 0.01 (2 on disk) |
+| DeepSpeed ZeRO-Inference | 9.28 (16 on CPU)  | 0.60 (4 on CPU) | 0.01 (1 on disk) |
+| Petals\*                 | -     | -    | 0.05 |
+| FlexGen                  | 25.26 (2 on GPU) | 7.32 (144 on CPU) | 0.69 (256 on disk) |
+| FlexGen with Compression | **29.12** (72 on GPU) | **8.38** (512 on CPU) | **1.12** (144 on CPU) |
 
 - Hardware: an NVIDIA T4 (16GB) instance on GCP with 208GB of DRAM and 1.5TB of SSD.  
 - Workload: input sequence length = 512, output sequence length = 32. The batch size is tuned to **a large value** that maximizes the generation throughput for each system.