vllm
llama.cpp
vllm | llama.cpp | |
---|---|---|
31 | 778 | |
19,344 | 57,984 | |
12.6% | - | |
9.9 | 10.0 | |
1 day ago | 2 days ago | |
Python | C++ | |
Apache License 2.0 | MIT License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
vllm
-
AI leaderboards are no longer useful. It's time to switch to Pareto curves
I guess the root cause of my claim is that OpenAI won't tell us whether or not GPT-3.5 is an MoE model, and I assumed it wasn't. Since GPT-3.5 is clearly nondeterministic at temp=0, I believed the nondeterminism was due to FPU stuff, and this effect was amplified with GPT-4's MoE. But if GPT-3.5 is also MoE then that's just wrong.
What makes this especially tricky is that small models are truly 100% deterministic at temp=0 because the relative likelihoods are too coarse for FPU issues to be a factor. I had thought 3.5 was big enough that some of its token probabilities were too fine-grained for the FPU. But that's probably wrong.
On the other hand, it's not just GPT, there are currently floating-point difficulties in vllm which significantly affect the determinism of any model run on it: https://github.com/vllm-project/vllm/issues/966 Note that a suggested fix is upcasting to float32. So it's possible that GPT-3.5 is using an especially low-precision float and introducing nondeterminism by saving money on compute costs.
Sadly I do not have the money[1] to actually run a test to falsify any of this. It seems like this would be a good little research project.
[1] Or the time, or the motivation :) But this stuff is expensive.
-
Mistral AI Launches New 8x22B Moe Model
The easiest is to use vllm (https://github.com/vllm-project/vllm) to run it on a Couple of A100's, and you can benchmark this using this library (https://github.com/EleutherAI/lm-evaluation-harness)
- FLaNK AI for 11 March 2024
-
Show HN: We got fine-tuning Mistral-7B to not suck
Great question! scheduling workloads onto GPUs in a way where VRAM is being utilised efficiently was quite the challenge.
What we found was the IO latency for loading model weights into VRAM will kill responsiveness if you don't "re-use" sessions (i.e. where the model weights remain loaded and you run multiple inference sessions over the same loaded weights).
Obviously projects like https://github.com/vllm-project/vllm exist but we needed to build out a scheduler that can run a fleet of GPUs for a matrix of text/image vs inference/finetune sessions.
disclaimer: I work on Helix
-
Mistral CEO confirms 'leak' of new open source AI model nearing GPT4 performance
FYI, vLLM also just added experiment multi-lora support: https://github.com/vllm-project/vllm/releases/tag/v0.3.0
Also check out the new prefix caching, I see huge potential for batch processing purposes there!
- VLLM Sacrifices Accuracy for Speed
- Easy, fast, and cheap LLM serving for everyone
- vllm
- Mixtral Expert Parallelism
- Mixtral 8x7B Support
llama.cpp
-
IBM Granite: A Family of Open Foundation Models for Code Intelligence
if you can compile stuff, then looking at llama.cpp (what ollama uses) is also interesting: https://github.com/ggerganov/llama.cpp
the server is here: https://github.com/ggerganov/llama.cpp/tree/master/examples/...
And you can search for any GGUF on huggingface
-
Ask HN: Affordable hardware for running local large language models?
Yes, Metal seems to allow a maximum of 1/2 of the RAM for one process, and 3/4 of the RAM allocated to the GPU overall. There’s a kernel hack to fix it, but that comes with the usual system integrity caveats. https://github.com/ggerganov/llama.cpp/discussions/2182
- Xmake: A modern C/C++ build tool
-
Better and Faster Large Language Models via Multi-Token Prediction
For anyone interested in exploring this, llama.cpp has an example implementation here:
https://github.com/ggerganov/llama.cpp/tree/master/examples/...
- Llama.cpp Bfloat16 Support
-
Fine-tune your first large language model (LLM) with LoRA, llama.cpp, and KitOps in 5 easy steps
Getting started with LLMs can be intimidating. In this tutorial we will show you how to fine-tune a large language model using LoRA, facilitated by tools like llama.cpp and KitOps.
- GGML Flash Attention support merged into llama.cpp
-
Phi-3 Weights Released
well https://github.com/ggerganov/llama.cpp/issues/6849
- Lossless Acceleration of LLM via Adaptive N-Gram Parallel Decoding
- Llama.cpp Working on Support for Llama3
What are some alternatives?
TensorRT - NVIDIA® TensorRT™ is an SDK for high-performance deep learning inference on NVIDIA GPUs. This repository contains the open source components of TensorRT.
ollama - Get up and running with Llama 3, Mistral, Gemma, and other large language models.
CTranslate2 - Fast inference engine for Transformer models
gpt4all - gpt4all: run open-source LLMs anywhere
lmdeploy - LMDeploy is a toolkit for compressing, deploying, and serving LLMs.
text-generation-webui - A Gradio web UI for Large Language Models. Supports transformers, GPTQ, AWQ, EXL2, llama.cpp (GGUF), Llama models.
Llama-2-Onnx
GPTQ-for-LLaMa - 4 bits quantization of LLaMA using GPTQ
tritony - Tiny configuration for Triton Inference Server
ggml - Tensor library for machine learning
faster-whisper - Faster Whisper transcription with CTranslate2
alpaca.cpp - Locally run an Instruction-Tuned Chat-Style LLM