fzysingularity

13 ms·

121.
Serving LLMs on a Budget(docs.nos.run)
2 points by fzysingularity 3y ago|0 comments
122.
by fzysingularity 3y ago
Classic knowledge distillation! I’d even argue that we won’t need 8x7b for fine-tuning here. Soon enough, phi-2 or phixtral models will be sufficiently powerful after fine tuning for these domains.
123.
by fzysingularity 3y ago
The war on price of serving LLMs - https://x.com/dylan522p/status/1736860043947937880?s=46&t=VJ...
124.
2024 predictions on the open-source LLM wars(twitter.com)
2 points by fzysingularity 3y ago|2 comments
125.
by fzysingularity 3y ago
This is cool! How much does a single song of ~1 min cost/take to generate?
126.
Fly.io for Multi-Cloud?
1 points by fzysingularity 3y ago|1 comments
127.
The AI Stack for the 3rd Epoch of Computing(spillai.substack.com)
4 points by fzysingularity 3y ago|0 comments
128.
My Own AI Server Cluster(erichartford.com)
3 points by fzysingularity 3y ago|0 comments
129.
by fzysingularity 3y ago
Here’s a fun project I hacked together last weekend — https://github.com/spillai/agi-pack TL;DR agi-pack is a Dockerfile generator for machine learning (ML) developers that is simple, hackable and extensible. Fun fact:
130.
AGI-pack: Dockerfile generator for ML developers(github.com)
1 points by fzysingularity 3y ago|1 comments
131.
by fzysingularity 3y ago
Here’s a fun project I hacked together this weekend. https://github.com/spillai/agi-pack TL;DR `agi-pack` is a Dockerfile generator for machine learning (ML) developers that is simple, hackable and extensible. Inspired
132.
by fzysingularity 3y ago
Also, isn't the author Tri Dao at Together AI now as their chief scientist?
133.
Blog – Introducing Intrinsic Flowstate – Intrinsic(intrinsic.ai)
1 points by fzysingularity 3y ago|0 comments
134.
Building LLM Applications for Production(huyenchip.com)
2 points by fzysingularity 3y ago|0 comments
135.
by fzysingularity 4y ago
Looks interesting and useful. How much data do you typically handle? And I imagine customers work with other DBs and stores alongside Baseplate? BTW your discord link is no longer active on the landing page. Mind sharing a new link?
136.
by fzysingularity 4y ago
Interesting - is there a good reference to back this claim? Curious to hear what overheads Faiss would have if it's configured with similar parameters to build the HNSW graphs. Is that what you A/B-tested in practice?
137.
by fzysingularity 4y ago
Big fan of Faiss - I've tried using several others (milvus, weaviate, opensearch, etc) but none struck the usability and configurability chord as much as Faiss did. I especially like their index-factory models. Once you figure out how
138.
by fzysingularity 4y ago
Thanks for your reply. - I have a good sense of the overheads - I was more curious about the latencies (ms) you are observing with the system today. - Out of curiosity, why did you pick Redis? Is it mostly due to familiarity and experience
139.
Chroma(trychroma.com)
1 points by fzysingularity 4y ago|0 comments
140.
by fzysingularity 4y ago
Congrats on the launch! Few questions/thoughts: - What kind of overheads do you have right now with calling this API? - What scales have you pressure-tested this with? Demo seems to show few 100s of embeddings. Selfishly, I'd lik
141.
OpenAI ChatGPT: Service Unavailable(chat.openai.com)
3 points by fzysingularity 4y ago|0 comments
142.
by fzysingularity 4y ago
What's next? Dreamfusion Video = Imagen Video (this) + Dreamfusion ( https://dreamfusion3d.github.io/ ) Fundamentally, I think we have all the pieces based on this work and Dreamfusion to make it work. From the looks of
143.
Benchmarking AWS Inferentia (inf1) chip performance(twitter.com)
2 points by fzysingularity 5y ago|0 comments
144.
by fzysingularity 5y ago
Out of curiosity, what kind of performance do you get with 100M rows with pinecone? Looking at your pricing tiers, ~100M rows would need ~200GB memory, and @ $0.1 / GB / hr that's $20 / hr if I'm not mistaken? Also,
145.
by fzysingularity 5y ago
Others have pointed this out already, but have a look at Milvus ( https://github.com/milvus-io/milvus ). I was able to get a simple version of it running with searches over ~3B vectors in under a second on a single machi
146.
by fzysingularity 6y ago
Toyota Research Institute (TRI) | Full-Time | ONSITE (covid WFH) | Bay Area TL;DR We're looking to grow our Machine Learning Engineering team at Toyota Research Institute (TRI). If you’re interested in developing bleeding-edge ML for o
147.
OpenCV NumPy Converter Using Boost::Python(github.com)
2 points by fzysingularity 12y ago|0 comments
148.
Learning Articulated Motions from Visual Demonstration(people.csail.mit.edu)
3 points by fzysingularity 12y ago|0 comments
149.
Bitcoin Transaction Graph Analysis(css.csail.mit.edu)
1 points by fzysingularity 13y ago|0 comments