card: restore Model tree (base_model) + Parameters table

Founder: fix all model cards with model tree and parameters. 2026-08-03 20:51:10 UTC+07:00. Packs: base_model for Hub tree findability only (not a finetune claim). GGUF: quantized relation. Parameters table = size surface (sidebar badge empty without weights).

README.md

@@ -1,6 +1,8 @@
 ---
 license: apache-2.0
 license_link: https://www.apache.org/licenses/LICENSE-2.0
+pipeline_tag: text-generation
+base_model: Qwen/Qwen3-Coder-Next
 language:
   - en
 tags:
@@ -22,34 +24,21 @@ tags:
 
 # Qwen3-Coder-Next-Spark-Agentic
 
-## Parameters (upstream model, not this pack)
-
-| Field | Value |
-|---|---|
-| Total parameters | **80B** |
-| Active parameters per token | **3B** |
-| Architecture | MoE · `Qwen3NextForCausalLM` |
-| Upstream native context | **262,144 tokens** |
-| This launch profile | **32,768-token configured ceiling; effective context unmeasured** |
-| Day-0 artifact | Official `Qwen3-Coder-Next-FP8` |
-| This repository | Deployment/evaluation pack only; no weights |
-| Specification source | [`Qwen/Qwen3-Coder-Next`](https://huggingface.co/Qwen/Qwen3-Coder-Next) |
-
-The parameter and native-context figures are upstream specifications. The blank
-Hub Parameters badge is expected because this repository contains no checkpoint.
-
-An independent, reproducible deployment pack for running the official
-[Qwen3-Coder-Next-FP8](https://huggingface.co/Qwen/Qwen3-Coder-Next-FP8) on one
-128 GB NVIDIA DGX Spark.
-
-> **Day-0 class: Preview.** This repository contains launch scripts, exact pins,
-> configuration, and evaluation scaffolding—not model weights and not a new
-> quantization. It makes no ship, hero, winner, throughput, or agent-reliability claim.
-
-> **Freeze status:** Source **freeze-attested** (2026-08-02 freeze-open PASS +
-> Steps 3–6 public tips). Class remains **Preview**. Exact DGX Spark load/tool/
-> throughput/reliability evidence is still pending. Must not be labeled
-> `runnable_day0` or hero.
+## Parameters
+
+| | |
+|--|--|
+| **Total** | **~80B** (HF count 79.67B BF16; non-embedding ~79B) |
+| **Active / token** | **~3B** |
+| **Arch** | MoE · `Qwen3NextForCausalLM` · 512 experts · 10 act · 1 shared |
+| **Upstream native context** | **262,144** tokens |
+| **This launch profile** | **32,768**-token configured ceiling; effective context unmeasured |
+| **Day-0 artifact** | Official `Qwen3-Coder-Next-FP8` |
+| **Source** | [Qwen/Qwen3-Coder-Next](https://huggingface.co/Qwen/Qwen3-Coder-Next) |
+| **This pack** | **deployment / evaluation pack only** — no weights here |
+
+> HF native **Parameters** badge stays empty (no checkpoint in this repo) — size is the table above.
+> Native **Model tree** is enabled via YAML `base_model: Qwen/Qwen3-Coder-Next` for findability only — this is **not** a finetune/adapter/quant of that checkpoint.
 
 ## What is ready