Model card: point to the repo for ranking thresholds

Model card: badges (Habr article, code, demo, benchmark, Colab, vLLM, license)

Add decision-model tag (#3)

Card: LLM baselines on razvilka

Model card: v0.4.0, MLX on Apple Silicon (#2)

Card: longer texts wording

Card: v0.3.0, async API (AsyncJudge, VllmJudge)

Card: raising the length limits

Card: input length

Card: credit УЭСМО, SberAI; citation

Card: v0.2.0, vLLM server section

Model card: vLLM server (frida-decisions 0.2.0) (#1)

Card: install v0.1.1

Card: peak GPU memory wording

Card: links to code, demo, benchmark and Colab

Model card: Colab links follow main

Model card: Colab links

Model card: numbers of the released artifacts

Model card: pin code to v0.1.0

FRIDA-Decisions: bf16 weights, head, int8 ONNX, model card