README.md: Correct heading

README.md

@@ -96,7 +96,7 @@ Finetune and evaluate [Mistral 7B Instruct v0.3](https://huggingface.co/mistrala
 |  | 32 | 4 | 83.6 samples/s | 2.6 samples/s (50%) | 11.4 GB |
 |  | 64 | 8 | 149.3 samples/s | 2.3 samples/s (44%) | 12.4 GB |
 
-### [mistral7b-bnb](mistral7b-bnb) multi GPU training with FSDP
+### [llama3.1-70b-bnb](llama3.1-70b-bnb) multi GPU training with FSDP
 
 Finetune and evaluate [Llama 3.1 70B Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct) with 4-bit [bitsandbytes quantisation](https://huggingface.co/docs/bitsandbytes/index) on the [MedMCQA](https://medmcqa.github.io) dataset on multiple GPUs on a single node using the [fully sharded data parallel (FSDP)](https://pytorch.org/tutorials/intermediate/FSDP_tutorial.html) approach.