cheesyFish

7 ms·

1.
Show HN: ExtractBench, an open-source schema extraction benchmark(github.com)
6 points by cheesyFish 1mo ago|0 comments
2.
by cheesyFish 5mo ago
I hope liteparse is working well for you! Funny enough since it doesn't parse as markdown, it can't be benchmarked with ParseBench (that tool is targeting slightly different use case with real-time agents etc.)
3.
Show HN: Local Document Parsing for Agents(llamaindex.ai)
20 points by cheesyFish 6mo ago|1 comments
4.
by cheesyFish 1y ago
yea good callout -- python workflows came first, and while we could have directly translated these, the ergonomics around classes in python are not exactly what JS/TS devs expect. So instead, the goal was to capture the spirit of event
5.
by cheesyFish 1y ago
We used pants initially (which I believe is similar to bazel). And indeed the dependency graphing it does was very helpful, but other parts of the tool motivated us to create something more bespoke and debuggable (we were only using like 20
6.
by cheesyFish 1y ago
Yea this definitely makes sense for smaller monorepos. For us, we ended up writing our own dependency graph parser to figure out what tests to run (which is easy enough with a single language like python honestly)
7.
by cheesyFish 1y ago
Yes! For example, previously with pants, users would hit a lot of weird errors since how tests run with pants is different than running tests locally with pytest We did not expect users to learn pants, but this often meant a lot of back and
8.
by cheesyFish 1y ago
I agree! I hope uv introduces more tools for monorepos or refines the workspaces concept. I saw workspaces require all dependencies to agree with eachother, which isn't quite possible in our repo
9.
Python Tooling at Scale: LlamaIndex’s Monorepo Overhaul(llamaindex.ai)
38 points by cheesyFish 1y ago|11 comments
10.
by cheesyFish 2y ago
It being a queue is one part of it yes. But the key is trying to provide tight integrations and take advantage of agentic features. Stuff like the orchestrator, having an external service to execute tools, etc.
11.
by cheesyFish 2y ago
I agree on the importance of letting the user have access to state! Right now there is actually the option for human in the loop. Additionally, I'd love to expand the monitor app a bit more to allow pausing, stepwise, rewind, etc.
12.
by cheesyFish 2y ago
It's definitely tough today, but its just a matter of a) using a smart LLM b) scoping down individual agents to a manageable set of actions As more LLMs come from companies and open-source, their reasoning abilities are only going to
13.
by cheesyFish 2y ago
maybe agent micro-services is a better way to frame it ha
14.
by cheesyFish 2y ago
Not quite. More like a framework to make LLMs/agents easier to deploy in a distributed fashion. We have a PR that shows how to deploy this with k8s!
15.
by cheesyFish 2y ago
Hey guys, Logan here! I've been busy building this for the past three weeks with the llama-index team. While it's still early days, I really think the agents-as-a-service vision is something worth building for. We have a solid set
16.
Show HN: Agents as Micro-Services(github.com)
6 points by cheesyFish 2y ago|0 comments
17.
by cheesyFish 2y ago
RRF is alright, but I've had better results with relative score, or distribution-based scoring. LlamaIndex has a module for exactly this https://docs.llamaindex.ai/en/stable/examples/retrievers/rel..