Open source Gemini alternatives

ECC matters

“> You're making a claim My claims are common sense and supported by the real life: regular people don't know what ECC is, and those who do find the problem's impact is too minor to get palpable benefits from fixing it. Why are you being pedantic if you aren't actually going to…”

ECC matters

“> My claims are common sense and supported by the real life: regular people don't know what ECC is, and those who do find the problem's impact is too minor to get palpable benefits from fixing it. Why are you being pedantic if you aren't actually going to bring arguments at the…”

Tracking: one gateway owns every local session (#106742) — post-merge fixes

“Tracking issue for [#106742]( (`feat/unified-gateway-runtime`). **#106742 merged to `main` on Oct 14** (merge commit `eb75627f`, branch head `2c01eb59`). This issue now tracks the post-merge fixes. Goal: every surface that talks to a **local** backend (CLI, TUI, locally-spawned…”

[skills-index-watchdog] Skills index is stale or degraded (degraded)

“Automated freshness probe failed. **Status:** `degraded` **Detail:** Index is 29.8h old (limit 26h) The Skills Hub at /docs/skills depends on `/docs/api/skills-index.json`. The unified index is rebuilt by `.github/workflows/skills-index.yml` (cron 6/18 UTC) and…”

The local LLM ecosystem doesn’t need Ollama

“The attribution and lock-in arguments are the loud parts of this story, but the quieter production reason to move is concurrency. llama.cpp's server takes parallel N with cont-batching enabled by default, which interleaves tokens from multiple requests inside a single batch and…”

The local LLM ecosystem doesn’t need Ollama

“Just tried llama.cpp NO, it is not simpler or even as simple as Ollama. There are multiple options-- llama server and cli, its not obivous which model to use. With ollama, its one file. And you get the models from their site, you can browse an easy list. I dont have the time to…”

Ponytail – make your AI agent think like the laziest senior dev in the room

“My thoughts exactly. The whole thing is essentially just these rules, and a metric ton of boilerplate for specific plugin systems. My own personal ponytail says this could just be this in a code block of a README”

Ponytail: Lazy Senior Engineer Skill

“I like how this repo has 159 files with 11635 LoC, and the load-bearing (!!!) part of it is a couple lines of natural language instructions: 1. *Does this need to exist at all?* Speculative need = skip it, say so in one line. (YAGNI) 2. *Already in this codebase?* A helper,…”

OSS Alternative to Open WebUI – ChatGPT-Like UI, API and CLI

“The dev has claimed that they are some sort of mission where their LLM gui is going to be responsible for saving humanity with them at the helm. I don't think they are completely rational. "No matter where humans go—on Earth or across the solar system—independence will be our…”

OSS Alternative to Open WebUI – ChatGPT-Like UI, API and CLI

“If we had a society of actual engineers and hackers, sure. However, LLMs for the vast majority of people are simple chatbot oracles. The people paying exorbinant wealth in investments are aiming for essentially the Apple of AI where it magically just works and creates a new…”

Show HN: Graphicle, a Rust port of Graphify with memory for coding agents

“Hey HN, I'm Vikram, I run KarLex AI. Right now if you want your coding agent to understand your code and remember things, you end up stitching tools together. Graphify for the code graph. Mem0 or Hindsight for memory. Each one runs separately, stores its own data, and you wire…”

Show HN: Graphicle, a Rust port of Graphify with memory for coding agents

“I read through the skill file of graphify and ran some of the commands it was telling the agent to run. The vibe I got is there was a lot of structured information returned that the model (and harness) isn't set up to work with effectively. It seems like a really interesting…”

Llama.cpp 30B runs with only 6GB of RAM now

“It's just the same llama.cpp repo everyone else is using. You just git clone it to your android phone in termux and then run make and you're done. Assuming you have the model file downloaded (you can use wget to download it) these are the instructions to install and run: pkg…”

Llama.cpp 30B runs with only 6GB of RAM now

“You can change the number of threads llama.cpp uses with the -t argument. By default it only uses 4. For example, if your CPU has 16 physical cores then you can run ./main -m model.bin -t 16 16 cores would be about 4x faster than the default 4 cores. Eventually you hit memory…”

Caveman: Why use many token when few token do trick

“As very much an outsider and, to some extent, apostate to all this, it's pretty astonishing to see. Unironically not just delegating all thinking to a sketchy and untrustworthy machine, but doubling down on it by aping the caveman in the belief that this will more effectively…”

Caveman: Why use many token when few token do trick

“That was my point. You (and I) tend to read verobose text and not argue with it, our brains are spoonfed reasoning chains and they seem to make sense. Caveman breaks that, so we have to actually think, there is no "thinking" done for us”

Capitalism is a paperclip maximizer

“You are right that the analogy breaks down at the assumption that humans would allow the paperclip machine to run amok; human wants do powerfully change and influence the market machine. At the same time, the abyss gazes into us, and human wants change in response to the machine…”

Capitalism is a paperclip maximizer

“I believe our divergence is somewhere under the definition of 'intelligence'. The word is very ill-defined, to the point that we are reduced to describing it functionally in many cases, as with the Turing test. >Human intelligence is the intellectual capacity of humans, which is…”

vLLM: Easy, Fast, and Cheap LLM Serving with PagedAttention

“Yes, vLLM focuses on maximizing throughput when the VRAM is fully utilized. Nevertheless, I believe users can still benefit from vLLM even if they don't utilize the memory to its full capacity, because vLLM also includes other optimizations orthogonal to the PagedAttention…”

vLLM: Easy, Fast, and Cheap LLM Serving with PagedAttention

“This really depends on what GPUs you use. If you GPUs has very small amount of memory, vLLM will help more. vLLM addresses the memory bottleneck for saving KV caches and hence increases the throughput.”

Pls help, stuck with AutoGGUF

“I tried to make ggufs of different models (one that was already available and one which I made using the lazymergekit). I always get the same error how ever. It's this one (I edited the model name out but it happens with both I tested. They are Mistral 7b based ones): ```…”

LazyMergeKit - Tensor model.final_layernorm.weight required but not present in model ...

“Hi there I'm trying to merge Phi-2 models using the following config: `MODEL_NAME = "..." yaml_config = """ models: - model: microsoft/phi-2 # no parameters necessary for base model - model: rhysjones/phi-2-orange parameters: density: 0.5 weight: 0.5 - model:…”

Error generating code. Check the Developer Console AND the backend logs for details. Feel free to open a Github issue.

“After clicking "update", two Error generating code were prompted. Check the Developer Console AND the backend logs for details. Feel free to open a Github issue. 1. I confirmed that my apikey can use gpt-4-vision-preview 2. I ran it through docker-compose and first ran echo…”

RUN ERROR OF MAC OS ?????

“VITE v4.5.0 ready in 1254 ms ➜ Local: 14:06:21 ➜ Network: use --host to expose 14:06:21 ➜ press h to show help 14:06:21 ERROR 14:06:25 [TypeScript] Found 0 errors. Watching for file changes.”

Why has the entire document changed

“The other day when I was reading it, the explanation of the catalog was still like this ![Image](”

Unsloth Studio

“From the README at : "Unsloth uses a dual-licensing model of Apache 2.0 and AGPL-3.0. The core Unsloth package remains licensed under Apache 2.0, while certain optional components, such as the Unsloth Studio UI are licensed under AGPL-3.0."”

Unsloth Studio

“LM Studio user here. Unsloth looks great, wanted to check it out, but there is no app file to download and install? Sorry I'm not familiar with the command line.”

The Max Headroom Incident

“Typically art is not directed at unsuspecting bystanders and rarely breaks laws that carry potential prison sentences[0]. Even so, I would say that artist understand that they are fucking with people, but they believe that the value of their expression outweighs it. Weather I…”

The Max Headroom Incident

“The Art of Noise with Max Headroom - Paranoimia (Official Video) Coca-Cola Max Headroom | Max Headroom Coke commercials Max Headroom on Sesame Street The Headroom Collection on Letterman, 1986-1990 Terry Wogan interviews Max Headroom On Max Headroom: The Most Misunderstood Joke…”