# ashhart/TensorFold

> LLM inference server for Apple Silicon (MLX) and NVIDIA GPUs with an OpenAI-compatible API and exact speculative decoding. Supports Nemotron, Qwen3.8, GLM-5.3 and Gemma 4 families with per-family kernels and draft models.

- Magnitude: 5.6 out of 10 — Early signal
- Stars: 534 total · +10 stars today, ≈ 58 by evening
- Star trust: star growth looks organic
- Category: Language models · Language: Python · License: MIT · Created: 2026-06-19 · Last push: 2026-09-28
- GitHub: https://github.com/ashhart/TensorFold · Page: https://gitnova.dev/en/r/ashhart/TensorFold

## Useful for

- Serve a local OpenAI-compatible endpoint on a Mac for 4-bit Qwen3.8-27B
- Pre-download a checkpoint and draft model with tensorfold pull before starting the server
- Verify drafted output matches serial output by comparing with draft: false

## Why it’s here

- 10 stars so far today, about 58 expected by the end of the day. The usual pace is 0 per day, so that's 58× as much.
- Before this, the repository barely got any stars — about 2 per day.
- The spike has held for 3 days in a row — not a one-off blip.
- GitHub Trending Python today: #4, +160 stars.
- About 16 forks a day — people are taking the code.

## Star trust

Star growth looks organic.

Star-trust labels are heuristics based on the repository’s behavior, not a check of every stargazer.

## Numbers

- Forks: 53
- Issues and pull requests: 67
- Watchers: 9
- Average over the last week: 78 per day
- Usual pace: 0 per day
- Stars in the last hour (measured): 4
- Latest release: v0.3.6.1 (2026-09-28)

## Stars per day, last 30 days (oldest → newest, today is partial)

2026-08-30 … 2026-09-28: 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 26, 252, 207, 10

## Spotted in now

- GitHub Trending Python today: #4, +160 stars

## Similar by description

1. **MiaAI-Lab/DeepSeek-v4.1-Flash-EXL3-2x-DGX-Sparks** — 0.5 · Steady · Language models · Python · +0 stars today, ≈ 0 by evening
   A local EXL3 2.9 bpw checkpoint of DeepSeek-V4.1-Flash (196 GiB, 39 shards) served via an OpenAI-compatible vLLM API on a 2x NVIDIA GB10 (DGX Spark) kit with tensor-parallel 2. Includes DSpark speculative decoding and memory tuning for…
   Full card: https://gitnova.dev/en/r/MiaAI-Lab/DeepSeek-v4.1-Flash-EXL3-2x-DGX-Sparks.md
2. **NVIDIA/TensorRT-LLM** — 1.6 · Steady · Language models · Python · +5 stars today, ≈ 9 by evening
   NVIDIA library for optimizing inference of large language models and visual generative models on GPUs, with a Python API, specialized kernels, and an efficient C++/Python runtime.
   Full card: https://gitnova.dev/en/r/NVIDIA/TensorRT-LLM.md
3. **NVIDIA/Model-Optimizer** — 4.3 · Peaking · Language models · Python · +17 stars today, ≈ 55 by evening
   NVIDIA library for compressing and accelerating models: quantization, pruning, distillation, NAS and speculative decoding with export to TensorRT-LLM, vLLM, SGLang. For ML engineers preparing models for deployment.
   Full card: https://gitnova.dev/en/r/NVIDIA/Model-Optimizer.md
4. **incoai/splash** — 2.5 · Cooling · Language models · C++ · +14 stars today, ≈ 32 by evening
   A local LLM inference engine for Apple silicon, specialized per model: serves OpenAI- and Anthropic-compatible APIs to coding agents on a single Mac.
   Full card: https://gitnova.dev/en/r/incoai/splash.md

---

Magnitude (0–10) measures how fast and how unusually interest in a repository is growing right now. It is not a quality score.
Days are UTC. “So far today” is a fact; “expected by the end of the day” is a forecast.
Summaries and use cases are written by an LLM (DeepSeek V4.1 Flash) from the README and may be inaccurate: verify specific claims (benchmarks, speed, hardware) in the repository itself.
Data as of 2026-09-28 13:02 UTC, updated every 30 minutes.