config.py

import os

# Model settings
LLM_MODEL = os.getenv("LLM_MODEL", "Qwen/Qwen2.5-0.5B-Instruct")
EMBEDDER_MODEL = os.getenv("EMBEDDER_MODEL", "sentence-transformers/all-MiniLM-L6-v2")

# Curated <32B open-weight causal LMs for local inference (shown in the UI dropdown).
AVAILABLE_MODELS = [
    "Qwen/Qwen2.5-0.5B-Instruct",
    "HuggingFaceTB/SmolLM2-135M-Instruct",
    "HuggingFaceTB/SmolLM2-360M-Instruct",
    "Qwen/Qwen2.5-1.5B-Instruct",
    "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
    "HuggingFaceTB/SmolLM2-1.7B-Instruct",
    "microsoft/Phi-3.5-mini-instruct",
]

MODEL_INFO = {
    "HuggingFaceTB/SmolLM2-135M-Instruct": (
        "⚡ **Fastest · 135M params** — Loads in ~10 s.  \n"
        "Good for exploring the interface. May produce looser compressions; "
        "switch up for high-fidelity output."
    ),
    "HuggingFaceTB/SmolLM2-360M-Instruct": (
        "🚀 **Fast · 360M params** — Loads in ~20 s.  \n"
        "Noticeably better compression than 135M with a small speed trade-off."
    ),
    "Qwen/Qwen2.5-0.5B-Instruct": (
        "⚡ **Fast · 500M params · Default** — Loads in ~25 s.  \n"
        "Strong instruction-following for its size; reliably respects token budgets. "
        "Best balance of speed and quality."
    ),
    "TinyLlama/TinyLlama-1.1B-Chat-v1.0": (
        "🚀 **Fast · 1.1B params** — Loads in ~40 s.  \n"
        "Fully open, no licence required. Good general-purpose compression at 1B scale."
    ),
    "Qwen/Qwen2.5-1.5B-Instruct": (
        "⚖️ **Balanced · 1.5B params** — Loads in ~60 s.  \n"
        "Solid all-rounder; strong semantic fidelity and budget adherence."
    ),
    "HuggingFaceTB/SmolLM2-1.7B-Instruct": (
        "⚖️ **Balanced · 1.7B params** — Loads in ~60 s.  \n"
        "Designed for edge inference; efficient on CPU."
    ),
    "microsoft/Phi-3.5-mini-instruct": (
        "🏆 **Best quality · 3.8B params** — Loads in ~2 min.  \n"
        "Strongest reasoning and fidelity in this list. GPU strongly recommended."
    ),
}

# Curated sentence-transformer embedding models for quality scoring.
AVAILABLE_EMBEDDER_MODELS = [
    "sentence-transformers/all-MiniLM-L6-v2",
    "sentence-transformers/all-mpnet-base-v2",
    "BAAI/bge-small-en-v1.5",
    "BAAI/bge-base-en-v1.5",
    "mixedbread-ai/mxbai-embed-large-v1",
    "Alibaba-NLP/gte-Qwen2-1.5B-instruct",
]

EMBEDDER_INFO = {
    "sentence-transformers/all-MiniLM-L6-v2": (
        "⚡ **Fast · 22M params · Default**  \n"
        "Great baseline. Scores are reliable for typical compression ratios. "
        "Runs comfortably on CPU — minimal overhead."
    ),
    "sentence-transformers/all-mpnet-base-v2": (
        "⚖️ **Balanced · 110M params**  \n"
        "Noticeably sharper quality scores than MiniLM, especially on longer texts. "
        "Small speed trade-off; fine on CPU."
    ),
    "BAAI/bge-small-en-v1.5": (
        "⚡ **Fast · 33M params**  \n"
        "Strong quality-to-size ratio — often matches MiniLM on accuracy while being "
        "slightly more sensitive to meaning shifts. Good CPU option."
    ),
    "BAAI/bge-base-en-v1.5": (
        "⚖️ **Balanced · 109M params**  \n"
        "Consistently strong on semantic similarity benchmarks. "
        "Scores will be more discriminating — small differences in compression quality show up more clearly."
    ),
    "mixedbread-ai/mxbai-embed-large-v1": (
        "🏆 **High quality · 335M params**  \n"
        "Top-tier similarity scores. Quality readings will be the most accurate here, "
        "but slower to load and run. GPU recommended."
    ),
    "Alibaba-NLP/gte-Qwen2-1.5B-instruct": (
        "🔬 **Best quality · 1.5B params**  \n"
        "Strongest semantic understanding in this list. Scores will reflect subtle meaning loss "
        "that smaller models miss. Requires significant RAM/VRAM — GPU strongly recommended."
    ),
}

# Compression settings
DEFAULT_TARGET_TOKENS = 500
MAX_NEW_TOKENS = 1024

# Gradio
APP_TITLE = "TinyPress"
SERVER_PORT = int(os.getenv("PORT", 7860))