TheBloke

/

Llama-2-70B-Chat-GPTQ

like 260

Hi everyone,

I'm trying to install and run Llama 3: 70B (140GB) on my system, which has dual RTX 4090 GPUs and 64GB of RAM. Despite the powerful hardware, I'm facing some issues due to the model's massive resource requirements.

For context, I was able to run smaller models like Llama 3: 8B-Instruct and Llama 2: 13B-chat-hf without any problems. However, I'm struggling with larger models like Llama 2: 70B-chat-hf and Llama 3: 70B-Instruct.

Has anyone managed to run Llama 3: 70B on a similar setup? If so, could you please share your experience and any tips?

Specifically, I'm looking for advice on:

                1. Whether it's feasible to run this model with my current hardware.
                2. If I need to upgrade my RAM or GPUs, what would you recommend?
                3. Alternative models that are easier to run on my existing setup.

Any help or guidance would be greatly appreciated! Thanks in advance.

@TheBloke