HowToRun_vLLM_Models Readme Fixes

[anirudTT]

Various bugs / improvements in the HowToRun_vLLM_Models README

• “official documentation” points to broken link and need to be fixed: https://github.com/tenstorrent/tt-studio/blob/main/HowToRun_vLLM_Models.md#4-run-the-setup-script-vllm-llama31-70b-only #128
• Table of contents is misaligned and some links to lead to the intended location: https://github.com/tenstorrent/tt-studio?tab=readme-ov-file#table-of-contents e.g. when this is clicked Running Llama3.1-70B in TT-Studio, it doesn’t go to that section #129
• Make Step 6. Copy the Environment File and Point to it in TT-Studio instructions more cleaner and clear #130
• Check formatting in example block (https://github.com/tenstorrent/tt-studio/blob/main/HowToRun_vLLM_Models.md#step-2-point-to-the-copied-environment-file) this might be missing double quotes etc. #131
• Fix duplicated https://github.com/tenstorrent/tt-studio/blob/main/HowToRun_vLLM_Models.md#step-2-update-the-tt-studio-environment-file #132
• User should set to some root directory: https://github.com/tenstorrent/tt-studio/blob/main/HowToRun_vLLM_Models.md#4-run-the-setup-script-vllm-llama31-70b-only, e.g. for me, it wasn’t tt-inference-server/tt_inference_server_persistent_volume but it was /home/USER/tt-inference-server/tt_inference_server_persistent_volume #133
• Structure needs to be revised: https://github.com/tenstorrent/tt-studio/blob/main/HowToRun_vLLM_Models.md#5-folder-structure-for-model-weights — the files are actually under /home/USER/tt-studio/tt_studio_persistent_volume/volume_id_tt-metal-llama-3.1-70b-instructv0.0.1/model_weights/repacked-llama-3.1-70b-instruct #134
• This is no longer valid: https://github.com/tenstorrent/tt-studio/blob/main/HowToRun_vLLM_Models.md#step-1-copy-the-environment-file, the path now changed to: /home/USER/tt-inference-server/persistent_volume/model_envs #135
• Explore if this bug happens even after the steps are done #137
• Llama 3.1 70B weights URL #138

> Process SpawnProcess-1:
> Traceback (most recent call last):
>   File "/tt-metal/ttnn/ttnn/operations/core.py", line 622, in as_tensor
>     tensor = ttnn._ttnn.tensor.load_tensor(cache_file_name, device=device)
> RuntimeError: Cannot open "/home/user/cache_root/tt_metal_cache/cache_repacked-llama-3.1-70b-instruct/layers.0.attention.wqkv_fused.weight_multi_device_8_dtype_BFLOAT8_B_layout_TILE.bin"