TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Show HN: LlamaGPT – Self-hosted, offline, private AI chatbot, powered by Llama 2

225 pointsby mayankchhabraalmost 2 years ago

12 comments

ccozanalmost 2 years ago
Ok, since is running all private, how can I add my own private data? For example I have a 20+ years of an email archive that I'd like to be ingested.
评论 #37155282 未加载
评论 #37154064 未加载
评论 #37154364 未加载
评论 #37155040 未加载
Atlas-Marblesalmost 2 years ago
Very cool, this looks like a combination of chatbot-ui and llama-cpp-python? A similar project I&#x27;ve been using is <a href="https:&#x2F;&#x2F;github.com&#x2F;serge-chat&#x2F;serge">https:&#x2F;&#x2F;github.com&#x2F;serge-chat&#x2F;serge</a>. Nous-Hermes-Llama2-13b is my daily driver and scores high on coding evaluations (<a href="https:&#x2F;&#x2F;huggingface.co&#x2F;spaces&#x2F;mike-ravkine&#x2F;can-ai-code-results" rel="nofollow noreferrer">https:&#x2F;&#x2F;huggingface.co&#x2F;spaces&#x2F;mike-ravkine&#x2F;can-ai-code-resul...</a>).
评论 #37156498 未加载
belvalalmost 2 years ago
Nice project! I could not find the information in the README.md, can I run this with a GPU? If so what do I need to change? Seems like it&#x27;s hardcoded to 0 in the run script: <a href="https:&#x2F;&#x2F;github.com&#x2F;getumbrel&#x2F;llama-gpt&#x2F;blob&#x2F;master&#x2F;api&#x2F;run.sh#L12">https:&#x2F;&#x2F;github.com&#x2F;getumbrel&#x2F;llama-gpt&#x2F;blob&#x2F;master&#x2F;api&#x2F;run.s...</a>
评论 #37164076 未加载
评论 #37154554 未加载
评论 #37153819 未加载
SubiculumCodealmost 2 years ago
I didn&#x27;t see any info on how this is different than installing&#x2F;running llamacpp or koboldcpp. New offerings are awesome of course, but what is it adding?
评论 #37154565 未加载
avivoalmost 2 years ago
What is the advantage of this versus running something like <a href="https:&#x2F;&#x2F;github.com&#x2F;simonw&#x2F;llm">https:&#x2F;&#x2F;github.com&#x2F;simonw&#x2F;llm</a> , which also gives you options to e.g. use <a href="https:&#x2F;&#x2F;github.com&#x2F;simonw&#x2F;llm-mlc">https:&#x2F;&#x2F;github.com&#x2F;simonw&#x2F;llm-mlc</a> for accelerated inference?
caesilalmost 2 years ago
So many projects still using GPT in their name.<p>Is the thinking here that OpenAI is not going to defend that trademark? Or just kicking the can down the road on rebranding until the C&amp;D letter arrives?
评论 #37153841 未加载
评论 #37153814 未加载
评论 #37153831 未加载
QuinnyPigalmost 2 years ago
I&#x27;ve been looking for something like this for a while. Nice!
stormfatheralmost 2 years ago
Which layers are best to use as vector embeddings? Is it the initial embedding layer afer tokenization? First hidden layer? Second?
synaesthesisxalmost 2 years ago
How this compare to just running llama.cpp locally?
评论 #37154530 未加载
albert_ealmost 2 years ago
Oh I thought this was a quick guide to host it on any server (AWS &#x2F; other clouds) of our choosing.
评论 #37154716 未加载
评论 #37150241 未加载
chasd00almost 2 years ago
is it a free model or is the politically-correct-only response constraints in place?
评论 #37154579 未加载
评论 #37153416 未加载
评论 #37153408 未加载
lazzlazzlazzalmost 2 years ago
(1) What are the best more creative&#x2F;less lobotomized versions of Llama 2? (2) What&#x27;s the best way to get one of those running in a similarly easy way?
评论 #37154220 未加载
评论 #37153869 未加载
评论 #37153851 未加载
评论 #37154651 未加载