Spaces:
Running
Running
| title: Smaug-Flash Explorer | |
| emoji: π | |
| colorFrom: indigo | |
| colorTo: pink | |
| sdk: static | |
| pinned: false | |
| license: mit | |
| short_description: Specs, benchmarks, real encoder, BYO-endpoint chat | |
| # π Smaug-Flash Explorer | |
| An interactive companion page for | |
| [`abacusai/Smaug-Flash`](https://huggingface.co/abacusai/Smaug-Flash) β Abacus.AI's agentic-coding | |
| fine-tune of [`deepseek-ai/DeepSeek-V4-Flash-0731`](https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731) | |
| (304.18 B params, 166.9 GB checkpoint across 48 shards, 1 M context, MIT). | |
| Everything runs **client-side in your browser**: no backend, no GPU, no quota. | |
| | Tab | What it actually does | | |
| |---|---| | |
| | π§ **Overview** | Spec table from `config.json` and the Hub's safetensors metadata, checkpoint anatomy, training approach (2Γ SFT + 1Γ KTO LoRA merged as full attention-only deltas), known behaviors and inference gotchas β plus live Hub stats fetched from the public API. | | |
| | π **Benchmarks** | The model card's numbers as Plotly charts: grouped Smaug-Flash vs base bars, a delta chart, and a LiveBench category radar β with the card's own figures embedded for comparison and its caveats spelled out. | | |
| | π€ **Prompt encoder** | The repo ships **no Jinja chat template**, so this tab runs the model's own `encoding/encoding_dsv4.py` β vendored here **byte-for-byte unmodified** and executed by Pyodide β to turn any message list into the exact prompt string the model was trained on. It also parses a raw completion back into OpenAI-format fields. | | |
| | π¬ **Playground** | A streaming chat client that speaks OpenAI-compatible SSE directly from your browser to an endpoint you choose: your own vLLM/SGLang deployment, a tunnelled remote one, or the HF router (for a plumbing check against a live Flash-class model). | | |
| | π **Deploy & size** | The card's vLLM and SGLang recipes, the reference `inference/` conversion path, a Python client snippet, and an interactive weight-memory calculator that maps quantization choices onto real GPU tiers and prices. | | |
| ## Why there is no in-Space inference | |
| The checkpoint is **304.18 B parameters / 166.9 GB on disk** (packed-FP4 experts + block-FP8 | |
| attention) and its Hub `inferenceProviderMapping` is **empty** β no serverless Inference Provider | |
| serves it. HF Spaces hardware tops out at ZeroGPU's 96 GB, and the native-format budget is ~165 GB of | |
| weights before any KV cache, so the weights cannot be loaded in a Space. This page therefore does the | |
| things that *do* work at zero GPU cost. | |
| To serve the model for real, see the **Deploy & size** tab β a dedicated multi-GPU tier for this | |
| model starts around $5.50/h. If you want a Space to serve it, a | |
| [community GPU grant](https://huggingface.co/docs/hub/spaces-gpus) is the route. | |
| ## Notes worth knowing | |
| * **Special-token delimiter.** The repo's `encoding_dsv4.py` *and* its `tokenizer.json` delimit | |
| special tokens with `U+FF5C` (`οΌΌ`) rather than ASCII `|` β near-identical in most fonts, which is | |
| why a hand-typed pipe silently fails to parse. The decoder repairs typed ASCII pipes and tells you | |
| when it did. | |
| * `reasoning_effort` β `low` (0-char preamble) / `high` (476 chars) / `max` (526 chars); `None` is | |
| treated as `low`, it only applies in `thinking` mode, and only on the first message. Budget up to | |
| 384 K output tokens for high/max. | |
| * The encoder is strict by design: malformed completions raise (`missing thinking`, `missing EOS | |
| token`, `Unexpected content at end`) instead of being patched up β the card is explicit about that. | |
| * Benchmark values in the charts are **digitized from the model card's figures**; the originals are | |
| embedded on the page and remain authoritative. | |
| ## Running it locally | |
| ```bash | |
| git clone https://huggingface.co/spaces/xdLocutus/Smaug-Flash-Explorer | |
| cd Smaug-Flash-Explorer | |
| python -m http.server 8765 # open http://127.0.0.1:8765 | |
| ``` | |
| Locally is also the easiest way to use the Playground against a local runtime | |
| (`http://127.0.0.1:8000/v1`): no TLS, no CORS friction. | |
| ## Optional: the Gradio variant | |
| `app.py` + `requirements.txt` in this repo are a **Gradio port of the same five tabs** (same data, | |
| same vendored encoder, plus an API/MCP surface). They are inert while the README frontmatter says | |
| `sdk: static`. To switch, replace the frontmatter with: | |
| ```yaml | |
| sdk: gradio | |
| sdk_version: 6.31.0 | |
| app_file: app.py | |
| python_version: "3.12" | |
| ``` | |
| That requires hardware β Gradio Spaces on a free account need PRO, and ZeroGPU additionally needs a | |
| 30-day-old account or an approved community grant. The Gradio app is designed for ZeroGPU: it keeps a | |
| no-op `@spaces.GPU` function so the runtime is attached but no quota is spent, since it never touches | |
| CUDA. | |
| ## Attribution | |
| Model card and repository contents (MIT) belong to Abacus.AI; `encoding_dsv4.py` is vendored | |
| unmodified from the model repo; the base architecture is DeepSeek-AI's. This Space is unofficial. | |