Chronos Rs GGUF

--- license: apache-2.0 library_name: gguf pipeline_tag: time-series-forecasting language: - en base_model: amazon/chronos-2 base_model_relation: quantized quantized_by: amaye15 tags: - gguf - time-series - forecasting - zero-shot - transformer - rust inference: false --- # Chronos-2 (Amazon) — GGUF GGUF conversion of [Amazon's Chronos-2](https://huggingface.co/amazon/chronos-2) — an encoder-only bidirectional time-series foundation model. Converted and run with [**zsfm**](https://github.com/amaye15/zsfm-rs), a Rust workspace that ports zero-shot forecasting and tabular foundation models to GGUF + [candle](https://github.com/huggingface/candle). No PyTorch, no Python runtime required to run inference. | F32 | F16 | Q8_0 | |---|---|---| | `chronos-f32.gguf` | `chronos-f16.gguf` | `chronos-q8.gguf` | F16 is generally the best size/accuracy trade-off; Q8_0 is smallest. This repo's default recommendation matches the upstream conversion default: **F16**. There's no `config.json` in this repo — GGUF embeds its own architecture metadata for the CLI, but the Python bindings still need `config.json` from [amazon/chronos-2](https://huggingface.co/amazon/chronos-2). **Context must be at least 32 timesteps** (Chronos-2's patch_size is 16; shorter contexts are rejected outright rather than padded — 32 gives headroom) — a shorter context fails with `context too short`. The examples below use a 32-value context. ## Use it ### Python (`pip install zsfm`) ```bash pip install zsfm huggingface_hub ``` ```python import zsfm from huggingface_hub import hf_hub_download gguf_path = hf_hub_download("amaye15/chronos-rs-gguf", "chronos-f16.gguf") config_path = hf_hub_download("amazon/chronos-2", "config.json") model = zsfm.ChronosModel(gguf_path, config_path) context = [0.85, 0.93, 1.01, 1.09, 1.17, 1.25, 1.33, 1.06, 1.14, 1.22, 1.3, 1.38, 1.46, 1.54, 1.27, 1.35, 1.43, 1.51, 1.59, 1.67, 1.75, 1.48, 1.56, 1.64, 1.72, 1.8, 1.88, 1.96, 1.69, 1.77, 1.85, 1.93] point = model.forecast(context, horizon=64) # -> List[float], the median (q0.5) forecast ``` Chronos has the richest Python surface of the single-checkpoint models here: besides `forecast()` (median), it also exposes `forecast_quantiles(context, horizon) -> List[List[float]]` (the full quantile matrix) and `quantiles() -> List[float]` (the quantile levels the matrix rows correspond to). ### Rust / CLI (`car...

--- license: mit library_name: gguf pipeline_tag: time-series-forecasting language: - en base_model: amazon/chronos-2 base_model_relation: quantized quantized_by: amaye15 tags: - gguf - time-series - forecasting - zero-shot - probabilistic - transformer - rust inference: false --- # chronos-rs Pure Rust converter and inference engine for [amazon/chronos-2](https://huggingface.co/amazon/chronos-2). Pre-converted GGUF files are available at [amaye15/chronos-rs-gguf](https://huggingface.co/amaye15/chronos-rs-gguf). Produces GGUF v3 files and runs native forecasting — no Python required. ## Build ```bash cargo build --release ``` ## Convert Downloads the model from HuggingFace and writes a GGUF file: ```bash # F16 (recommended — good precision/size trade-off) ./target/release/chronos-rs convert --model amazon/chronos-2 --dtype f16 --output gguf/chronos-f16.gguf # Q8_0 (smallest, ~2.8× compression vs F16) ./target/release/chronos-rs convert --dtype q8 --output gguf/chronos-q8.gguf # F32 (full precision) ./target/release/chronos-rs convert --dtype f32 --output gguf/chronos-f32.gguf ``` To convert all dtypes at once: ```bash ./scripts/convert_all.sh ``` HuggingFace token (optional for public models): ```bash HF_TOKEN=hf_... ./scripts/convert_all.sh ``` ## Inspect tensors Print all tensor names and shapes from a `.safetensors` checkpoint: ```bash ./target/release/chronos-rs inspect-tensors models/model.safetensors ``` ## Infer Run univariate quantile forecasting from a GGUF file: ```bash echo '{"context": [1.0, 1.2, 1.5, 1.3, 1.8, 2.0, 1.9, 2.1], "horizon": 64}' \ | ./target/release/chronos-rs infer \ --gguf gguf/chronos-f16.gguf \ --config models/config.json ``` Output is JSON in an OpenAI-compatible forecast format: ```json { "id": "forecast-000001932b7a1234", "object": "forecast", "created": 1749686400, "model": "chronos", "choices": [{ "index": 0, "forecast": { "point": [2.1, 2.3, 2.5, "..."], "quantiles": { "0.10": [1.8, 2.0, 2.2, "..."], "0.50": [2.1, 2.3, 2.5, "..."], "0.90": [2.4, 2.6, 2.8, "..."] } }, "finish_reason": "stop" }], "usage": {"context_length": 8, "forecast_length": 64} } ``` `point` is the median (q0.5) forecast; all quantile levels from `config.json` are included. **Batch inference** — pass multiple series as a nested array to get one `Choice` per series: ```bash echo '{"context": [[1.0,...

Source context: 124 downloads · 1 likes · Pipeline time-series-forecasting · Library gguf · Repo amaye15/chronos-rs-gguf