Alternatives
Products that do what GitHub does
BERT, Tokenizer, Python, WordPiece, pybind11,C++,Flash,Trie
- 1TA
TokenDagger is a drop-in replacement for OpenAI’s Tiktoken (the tokenizer behind Llama 3, Mistral, GPT-3.*, etc.). It’s written in C++ 17 with thin Python bindings, keeps the exact same BPE vocab/special-token rules, and focuses on raw speed. I’m teaching myself LLM internals by re-implementing the stack from first principles. Profiling TikToken’s Python/Rust implementation showed a lot of time was spent doing regex matching. Most of my perf gains come from a) using a faster jit-compiled regex engine; and b) simplifying the algorithm to forego regex matching special tokens at all.…
2025 · github.com
- 2F1
I built a tokenizer in C++ with a Python binding that outperforms HuggingFace tokenizers by 10x on large inputs. It's optimized for minimal memory usage and latency. Benchmarks and comparison included in README. Would love feedback or contributions!
2025 · github.com
- 3
Best-in-class intelligence for your high-volume workloads
Mar 2026 · blog.google
- 4WD
2022 · github.com
- 58F
Hi HN! I'm just sharing a project I've been working on during the LLM Efficiency Challenge - you can now finetune Llama with QLoRA 5x faster than Huggingface's original implementation on your own local GPU. Some highlights: 1. Manual autograd engine - hand derived backprop steps. 2. QLoRA / LoRA 80% faster, 50% less memory. 3. All kernels written in OpenAI's Triton language. 4. 0% loss in accuracy - no approximation methods - all exact. 5. No change of hardware necessary. Supports NVIDIA GPUs since 2018+. CUDA 7.5+. 6. Flash Attention support via Xformers. 7. Supports 4bit and 16bit…
2023 · github.com
- 6MP
Aug 2026 · deepgrove.ai
- 7FF
A few months ago, I benchmarked FastAPI on an i9 MacBook Pro. I couldn't believe my eyes. A primary REST endpoint to `sum` two integers took 6 milliseconds to evaluate. It is okay if you are targeting a server in another city, but it should be less when your client and server apps are running on the same machine. FastAPI would have bottleneck-ed the inference of our lightweight UForm neural networks recently trending on HN under the title "Beating OpenAI CLIP with 100x less data and compute". (Thank you all for the kind words!) So I wrote another library. It has been a while since I have…
2023 · github.com
- 8CA
I built Chonkie because I was tired of rewriting chunking code for RAG applications. Existing libraries were either too bloated (80MB+) or too basic, with no middle ground. Core features: - 21MB default install vs 80-171MB alternatives - 33x faster token chunking than popular alternatives - Supports multiple chunking strategies: token, word, sentence, and semantic - Works with all major tokenizers (transformers, tokenizers, tiktoken) - Zero external dependencies for basic functionality Technical optimizations: - Uses tiktoken with multi-threading for faster tokenization - Implements…
2024 · github.com
- 9
- 10LT
2023 · github.com
- 11RB
Sep 2025 · github.com
- 12
Spotify Wrapped for Claude, Codex & a Public leaderboard.
Jun 2026 · whoburnedmore.com
- 13OL
I've been working on Fast LiteLLM - a Rust acceleration layer for the popular LiteLLM library - and I had some interesting learnings that might resonate with other developers trying to squeeze performance out of existing systems. My assumption was that LiteLLM, being a Python library, would have plenty of low-hanging fruit for optimization. I set out to create a Rust layer using PyO3 to accelerate the performance-critical parts: token counting, routing, rate limiting, and connection pooling. The Approach - Built Rust implementations for token counting using tiktoken-rs - Added lock-free data…
Nov 2025 · github.com
- 14AA
Hey HN! We built Achilles, a tool that automatically accelerates your Python code. It identifies performance bottlenecks, rewrites those functions in optimized C++, and seamlessly patches them into your running program—without you changing a single line of code. In CPU-intensive, loop-heavy tasks, we've observed performance improvements of 100-1000x. Achilles can be installed via pip and works with just a single command. We'd appreciate your feedback, and feel free to give us a star if you find it interesting!
2025 · github.com
- 15
- 16FS
I want to share a really dumb, but very practical project I have packaged this summer, to perform operations on strings much faster. I was using Python to work with a multi-terabyte newline-delimited file. Reading, splitting, and shuffling it was a nightmare. So, I wrapped a trivial hardware-friendly heuristic I've been using for the last few years into a CPython library. The part I enjoyed the most is implementing SIMD behavior without SIMD instructions... Using 64-bit words to work at 8-bit granularity. Unlike conventional SIMD, the code would remain the same for ~~almost~~ any hardware.…
2023 · ashvardanian.com
- 17
I wanted to know how fast a 26B mixture-of-experts model could run on a desktop CPU with no GPU. Got ~40 tok/s single-stream (lossless) and ~124 batched. The surprising part was the byte budget: for this model you compress the output head (32% of per-token bytes), not the experts (16%). The writeup has the bandwidth roofline and the dead-ends; the repo has the reproducible recipe. Happy to answer questions. Repo: https://github.com/arun-prasath2005/gemma4-cpu-moe
Jun 2026 · apeg.dev
- 18
I use Claude Code, Codex and Cursor (and sometimes Antigravity) basically every day, and could never tell how much I was actually consuming across all of them. So I built TokenMaxxer. A small CLI reads the files these tools already write locally and puts it all in one dashboard, broken out by tool, model, provider and day. It covers 18 tools now, and you get a profile page with your daily activity, cost estimates, and your top models and tools. There's also a global leaderboard if you want to compete against other TokenMaxxers! I'd love to see if anyone can beat the first place (currently…
Aug 2026 · tokenmaxxer.xyz
- 19
- 20IE
Quick note on how it works and how I've done my batch embedding engine IgniteMS. The whole thing runs as one process using Rust, reading input, tokenizing, packing batches, keeping the queue full. TensorRT handles inference. Python is only as a wrapper. I built it this way because when you use more than couple of GPUs, the GPUs stop being the problem. CPU cannot feed them fast enough. One A100 can go through batches faster than Python can tokenize and feed, so the GPU just sits there idle waiting for work. Most of my time went into optimizing this. At 8 GPUs that was basically the entire…
Jun 2026 · github.com
- 21AC
2015 · github.com
- 22
Token-efficiency linter for LLM prompts and payloads - ritenv/tokensift
9d ago · github.com
- 23UF
2021 · github.com
- 24
I built a specialized package of DeepSeek V4 Flash 0731 (originally 284B total parameters, 13B active), preserving reasoning, tool calling and coding capabilities: https://huggingface.co/steadfastgaze/DeepSeek-V4-Flash-0731-... I let it write a minimal C compiler targeting ARM64, then test the result with Fibonacci and FizzBuzz programs, and it succeeded in less than 1 hour, with the full recording at: https://youtu.be/XiwSilmV8B0 You can run it on Silicon Macs with my engine https://github.com/steadfastgaze/MoEspresso, while one of the…
22d ago · huggingface.co
Ranked by how close each launch is in meaning, then by votes. Refine with a description →