title | date | layout | categories | tags | giscus_comments | authors | ||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
How to run LLM on a laptop? |
2024-09-20 |
post |
linux |
|
true |
|
{% highlight python %} Step 1: Get a laptop with >16 GB RAM and core i7 processor (GPU too is good, but I did not use it, It took 800% from cpu, 10% from Memory 4GB, for a HelloWorld) Step 2: Get Permission to LLAMA LLM usage from meta (Waiting Time 24 - 48 Hours, frankly unnecessary) Step 3: Install HuggingFace-hub Step 4: Download any LLAMA model variant from HF using huggingface-cli (7B) Step 5: Install ctransformers Step 6: Run the script (that you create) that opens the downloaded LLM file and use prompts on it (Check any HF LLAMA model for this) {% endhighlight %}
What do you think? Let me know your thoughts in the comments below.