app.py

import gradio as gr
import requests
import pandas as pd
import plotly.graph_objects as go
from datetime import datetime
import os

HF_TOKEN = os.getenv("HF_TOKEN")

target_models = {
    "openfree/flux-lora-korea-palace": "https://huggingface.co/openfree/flux-lora-korea-palace",
    "seawolf2357/hanbok": "https://huggingface.co/seawolf2357/hanbok",
    "seawolf2357/ntower": "https://huggingface.co/seawolf2357/ntower",

    "seawolf2357/flux-lora-military-artillery-k9": "https://huggingface.co/seawolf2357/flux-lora-military-artillery-k9",
    "openfree/claude-monet": "https://huggingface.co/openfree/claude-monet",
    "LGAI-EXAONE/EXAONE-3.5-32B-Instruct": "https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-32B-Instruct",
    "LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct": "https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct",
    "LGAI-EXAONE/EXAONE-3.5-7.8B-Instruct": "https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-7.8B-Instruct",
    "ginipick/flux-lora-eric-cat": "https://huggingface.co/ginipick/flux-lora-eric-cat",
    "seawolf2357/flux-lora-car-rolls-royce": "https://huggingface.co/seawolf2357/flux-lora-car-rolls-royce",
    "moreh/Llama-3-Motif-102B-Instruct": "https://huggingface.co/moreh/Llama-3-Motif-102B-Instruct",
    "OnomaAIResearch/Illustrious-xl-early-release-v0": "https://huggingface.co/OnomaAIResearch/Illustrious-xl-early-release-v0",
    "upstage/solar-pro-preview-instruct": "https://huggingface.co/upstage/solar-pro-preview-instruct",   
    "NCSOFT/VARCO-VISION-14B": "https://huggingface.co/NCSOFT/VARCO-VISION-14B",
    "NCSOFT/Llama-VARCO-8B-Instruct": "https://huggingface.co/NCSOFT/Llama-VARCO-8B-Instruct",
    "NCSOFT/VARCO-VISION-14B-HF": "https://huggingface.co/NCSOFT/VARCO-VISION-14B-HF",
    "KAERI-MLP/llama-3.1-Korean-AtomicGPT-Bllossom-8B": "https://huggingface.co/KAERI-MLP/llama-3.1-Korean-AtomicGPT-Bllossom-8B",
    "dnotitia/Llama-DNA-1.0-8B-Instruct": "https://huggingface.co/dnotitia/Llama-DNA-1.0-8B-Instruct",    
    "Bllossom/llama-3.2-Korean-Bllossom-3B": "https://huggingface.co/Bllossom/llama-3.2-Korean-Bllossom-3B",

    "unidocs/llama-3.1-8b-komedic-instruct": "https://huggingface.co/unidocs/llama-3.1-8b-komedic-instruct",
    "unidocs/llama-3.2-3b-komedic-instruct": "https://huggingface.co/unidocs/llama-3.2-3b-komedic-instruct",    
    "etri-lirs/eagle-3b-preview": "https://huggingface.co/etri-lirs/eagle-3b-preview",
    "kakaobrain/kogpt": "https://huggingface.co/kakaobrain/kogpt", 

    "Saxo/Linkbricks-Horizon-AI-Korean-Gemma-2-sft-dpo-27B": "https://huggingface.co/Saxo/Linkbricks-Horizon-AI-Korean-Gemma-2-sft-dpo-27B",
    "AALF/gemma-2-27b-it-SimPO-37K": "https://huggingface.co/AALF/gemma-2-27b-it-SimPO-37K",
    "nbeerbower/mistral-nemo-wissenschaft-12B": "https://huggingface.co/nbeerbower/mistral-nemo-wissenschaft-12B",
    "Saxo/Linkbricks-Horizon-AI-Korean-Mistral-Nemo-sft-dpo-12B": "https://huggingface.co/Saxo/Linkbricks-Horizon-AI-Korean-Mistral-Nemo-sft-dpo-12B",
    "princeton-nlp/gemma-2-9b-it-SimPO": "https://huggingface.co/princeton-nlp/gemma-2-9b-it-SimPO",
    "migtissera/Tess-v2.5-Gemma-2-27B-alpha": "https://huggingface.co/migtissera/Tess-v2.5-Gemma-2-27B-alpha",
    "DeepMount00/Llama-3.1-8b-Ita": "https://huggingface.co/DeepMount00/Llama-3.1-8b-Ita",
    "cognitivecomputations/dolphin-2.9.3-mistral-nemo-12b": "https://huggingface.co/cognitivecomputations/dolphin-2.9.3-mistral-nemo-12b",
    "ai-human-lab/EEVE-Korean_Instruct-10.8B-expo": "https://huggingface.co/ai-human-lab/EEVE-Korean_Instruct-10.8B-expo",
    "VAGOsolutions/Llama-3.1-SauerkrautLM-8b-Instruct": "https://huggingface.co/VAGOsolutions/Llama-3.1-SauerkrautLM-8b-Instruct",
    "Saxo/Linkbricks-Horizon-AI-Korean-llama-3.1-sft-dpo-8B": "https://huggingface.co/Saxo/Linkbricks-Horizon-AI-Korean-llama-3.1-sft-dpo-8B",
    "AIDX-ktds/ktdsbaseLM-v0.12-based-on-openchat3.5": "https://huggingface.co/AIDX-ktds/ktdsbaseLM-v0.12-based-on-openchat3.5",
    "mlabonne/Daredevil-8B-abliterated": "https://huggingface.co/mlabonne/Daredevil-8B-abliterated",
    "ENERGY-DRINK-LOVE/eeve_dpo-v3": "https://huggingface.co/ENERGY-DRINK-LOVE/eeve_dpo-v3",
    "migtissera/Trinity-2-Codestral-22B": "https://huggingface.co/migtissera/Trinity-2-Codestral-22B",
    "Saxo/Linkbricks-Horizon-AI-Korean-llama3.1-sft-rlhf-dpo-8B": "https://huggingface.co/Saxo/Linkbricks-Horizon-AI-Korean-llama3.1-sft-rlhf-dpo-8B",
    "mlabonne/Daredevil-8B-abliterated-dpomix": "https://huggingface.co/mlabonne/Daredevil-8B-abliterated-dpomix",
    "yanolja/EEVE-Korean-Instruct-10.8B-v1.0": "https://huggingface.co/yanolja/EEVE-Korean-Instruct-10.8B-v1.0",
    "vicgalle/Configurable-Llama-3.1-8B-Instruct": "https://huggingface.co/vicgalle/Configurable-Llama-3.1-8B-Instruct",
    "T3Q-LLM/T3Q-LLM1-sft1.0-dpo1.0": "https://huggingface.co/T3Q-LLM/T3Q-LLM1-sft1.0-dpo1.0",
    "Eurdem/Defne-llama3.1-8B": "https://huggingface.co/Eurdem/Defne-llama3.1-8B",
    "BAAI/Infinity-Instruct-7M-Gen-Llama3_1-8B": "https://huggingface.co/BAAI/Infinity-Instruct-7M-Gen-Llama3_1-8B",
    "BAAI/Infinity-Instruct-3M-0625-Llama3-8B": "https://huggingface.co/BAAI/Infinity-Instruct-3M-0625-Llama3-8B",
    "T3Q-LLM/T3Q-LLM-sft1.0-dpo1.0": "https://huggingface.co/T3Q-LLM/T3Q-LLM-sft1.0-dpo1.0",
    "BAAI/Infinity-Instruct-7M-0729-Llama3_1-8B": "https://huggingface.co/BAAI/Infinity-Instruct-7M-0729-Llama3_1-8B",
    "mightbe/EEVE-10.8B-Multiturn": "https://huggingface.co/mightbe/EEVE-10.8B-Multiturn",
    "hyemijo/omed-llama3.1-8b": "https://huggingface.co/hyemijo/omed-llama3.1-8b",
    "yanolja/Bookworm-10.7B-v0.4-DPO": "https://huggingface.co/yanolja/Bookworm-10.7B-v0.4-DPO",
    "algograp-Inc/algograpV4": "https://huggingface.co/algograp-Inc/algograpV4",
    "lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top75": "https://huggingface.co/lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top75",
    "chihoonlee10/T3Q-LLM-MG-DPO-v1.0": "https://huggingface.co/chihoonlee10/T3Q-LLM-MG-DPO-v1.0",
    "vicgalle/Configurable-Hermes-2-Pro-Llama-3-8B": "https://huggingface.co/vicgalle/Configurable-Hermes-2-Pro-Llama-3-8B",
    "RLHFlow/LLaMA3-iterative-DPO-final": "https://huggingface.co/RLHFlow/LLaMA3-iterative-DPO-final",
    "SEOKDONG/llama3.1_korean_v0.1_sft_by_aidx": "https://huggingface.co/SEOKDONG/llama3.1_korean_v0.1_sft_by_aidx",
    "spow12/Ko-Qwen2-7B-Instruct": "https://huggingface.co/spow12/Ko-Qwen2-7B-Instruct",
    "BAAI/Infinity-Instruct-3M-0625-Qwen2-7B": "https://huggingface.co/BAAI/Infinity-Instruct-3M-0625-Qwen2-7B",
    "lightblue/suzume-llama-3-8B-multilingual-orpo-borda-half": "https://huggingface.co/lightblue/suzume-llama-3-8B-multilingual-orpo-borda-half",
    "T3Q-LLM/T3Q-LLM1-CV-v2.0": "https://huggingface.co/T3Q-LLM/T3Q-LLM1-CV-v2.0",
    "migtissera/Trinity-2-Codestral-22B-v0.2": "https://huggingface.co/migtissera/Trinity-2-Codestral-22B-v0.2",
    "sinjy1203/EEVE-Korean-Instruct-10.8B-v1.0-Grade-Retrieval": "https://huggingface.co/sinjy1203/EEVE-Korean-Instruct-10.8B-v1.0-Grade-Retrieval",
    "MaziyarPanahi/Llama-3-8B-Instruct-v0.10": "https://huggingface.co/MaziyarPanahi/Llama-3-8B-Instruct-v0.10",
    "MaziyarPanahi/Llama-3-8B-Instruct-v0.9": "https://huggingface.co/MaziyarPanahi/Llama-3-8B-Instruct-v0.9",
    "zhengr/MixTAO-7Bx2-MoE-v8.1": "https://huggingface.co/zhengr/MixTAO-7Bx2-MoE-v8.1",
    "TIGER-Lab/MAmmoTH2-8B-Plus": "https://huggingface.co/TIGER-Lab/MAmmoTH2-8B-Plus",
    "OpenBuddy/openbuddy-qwen1.5-14b-v21.1-32k": "https://huggingface.co/OpenBuddy/openbuddy-qwen1.5-14b-v21.1-32k",
    "haoranxu/Llama-3-Instruct-8B-CPO-SimPO": "https://huggingface.co/haoranxu/Llama-3-Instruct-8B-CPO-SimPO",
    "Weyaxi/Einstein-v7-Qwen2-7B": "https://huggingface.co/Weyaxi/Einstein-v7-Qwen2-7B",
    "DKYoon/kosolar-hermes-test": "https://huggingface.co/DKYoon/kosolar-hermes-test",
    "vilm/Quyen-Pro-v0.1": "https://huggingface.co/vilm/Quyen-Pro-v0.1",
    "chihoonlee10/T3Q-LLM-MG-v1.0": "https://huggingface.co/chihoonlee10/T3Q-LLM-MG-v1.0",
    "lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top25": "https://huggingface.co/lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top25",
    "ai-human-lab/EEVE-Korean-10.8B-RAFT": "https://huggingface.co/ai-human-lab/EEVE-Korean-10.8B-RAFT",
    "princeton-nlp/Llama-3-Base-8B-SFT-RDPO": "https://huggingface.co/princeton-nlp/Llama-3-Base-8B-SFT-RDPO",
    "MaziyarPanahi/Llama-3-8B-Instruct-v0.8": "https://huggingface.co/MaziyarPanahi/Llama-3-8B-Instruct-v0.8",
    "chihoonlee10/T3Q-ko-solar-dpo-v7.0": "https://huggingface.co/chihoonlee10/T3Q-ko-solar-dpo-v7.0",
    "jondurbin/bagel-8b-v1.0": "https://huggingface.co/jondurbin/bagel-8b-v1.0",
    "DeepMount00/Llama-3-8b-Ita": "https://huggingface.co/DeepMount00/Llama-3-8b-Ita",
    "VAGOsolutions/Llama-3-SauerkrautLM-8b-Instruct": "https://huggingface.co/VAGOsolutions/Llama-3-SauerkrautLM-8b-Instruct",
    "princeton-nlp/Llama-3-Instruct-8B-ORPO-v0.2": "https://huggingface.co/princeton-nlp/Llama-3-Instruct-8B-ORPO-v0.2",
    "AIDX-ktds/ktdsbaseLM-v0.11-based-on-openchat3.5": "https://huggingface.co/AIDX-ktds/ktdsbaseLM-v0.11-based-on-openchat3.5",
    "princeton-nlp/Llama-3-Base-8B-SFT-KTO": "https://huggingface.co/princeton-nlp/Llama-3-Base-8B-SFT-KTO",
    "maywell/Mini_Synatra_SFT": "https://huggingface.co/maywell/Mini_Synatra_SFT",
    "princeton-nlp/Llama-3-Base-8B-SFT-ORPO": "https://huggingface.co/princeton-nlp/Llama-3-Base-8B-SFT-ORPO",
    "princeton-nlp/Llama-3-Instruct-8B-CPO-v0.2": "https://huggingface.co/princeton-nlp/Llama-3-Instruct-8B-CPO-v0.2",
    "spow12/Qwen2-7B-ko-Instruct-orpo-ver_2.0_wo_chat": "https://huggingface.co/spow12/Qwen2-7B-ko-Instruct-orpo-ver_2.0_wo_chat",
    "princeton-nlp/Llama-3-Base-8B-SFT-DPO": "https://huggingface.co/princeton-nlp/Llama-3-Base-8B-SFT-DPO",
    "princeton-nlp/Llama-3-Instruct-8B-ORPO": "https://huggingface.co/princeton-nlp/Llama-3-Instruct-8B-ORPO",
    "lcw99/llama-3-10b-it-kor-extented-chang": "https://huggingface.co/lcw99/llama-3-10b-it-kor-extented-chang",
    "migtissera/Llama-3-8B-Synthia-v3.5": "https://huggingface.co/migtissera/Llama-3-8B-Synthia-v3.5",
    "megastudyedu/M-SOLAR-10.7B-v1.4-dpo": "https://huggingface.co/megastudyedu/M-SOLAR-10.7B-v1.4-dpo",
    "T3Q-LLM/T3Q-LLM-solar10.8-sft-v1.0": "https://huggingface.co/T3Q-LLM/T3Q-LLM-solar10.8-sft-v1.0",
    "maywell/Synatra-10.7B-v0.4": "https://huggingface.co/maywell/Synatra-10.7B-v0.4",
    "nlpai-lab/KULLM3": "https://huggingface.co/nlpai-lab/KULLM3",
    "abacusai/Llama-3-Smaug-8B": "https://huggingface.co/abacusai/Llama-3-Smaug-8B",
    "gwonny/nox-solar-10.7b-v4-kolon-ITD-5-v2.1": "https://huggingface.co/gwonny/nox-solar-10.7b-v4-kolon-ITD-5-v2.1",
    "BAAI/Infinity-Instruct-3M-0625-Mistral-7B": "https://huggingface.co/BAAI/Infinity-Instruct-3M-0625-Mistral-7B",
    "openchat/openchat_3.5": "https://huggingface.co/openchat/openchat_3.5",
    "T3Q-LLM/T3Q-LLM1-v2.0": "https://huggingface.co/T3Q-LLM/T3Q-LLM1-v2.0",
    "T3Q-LLM/T3Q-LLM1-CV-v1.0": "https://huggingface.co/T3Q-LLM/T3Q-LLM1-CV-v1.0",
    "ONS-AI-RESEARCH/ONS-SOLAR-10.7B-v1.1": "https://huggingface.co/ONS-AI-RESEARCH/ONS-SOLAR-10.7B-v1.1",
    "macadeliccc/Samantha-Qwen-2-7B": "https://huggingface.co/macadeliccc/Samantha-Qwen-2-7B",
    "openchat/openchat-3.5-0106": "https://huggingface.co/openchat/openchat-3.5-0106",
    "NousResearch/Nous-Hermes-2-SOLAR-10.7B": "https://huggingface.co/NousResearch/Nous-Hermes-2-SOLAR-10.7B",
    "UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter1": "https://huggingface.co/UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter1",
    "MTSAIR/multi_verse_model": "https://huggingface.co/MTSAIR/multi_verse_model",
    "gwonny/nox-solar-10.7b-v4-kolon-ITD-5-v2.0": "https://huggingface.co/gwonny/nox-solar-10.7b-v4-kolon-ITD-5-v2.0",
    "VIRNECT/llama-3-Korean-8B": "https://huggingface.co/VIRNECT/llama-3-Korean-8B",
    "ENERGY-DRINK-LOVE/SOLAR_merge_DPOv3": "https://huggingface.co/ENERGY-DRINK-LOVE/SOLAR_merge_DPOv3",
    "SeaLLMs/SeaLLMs-v3-7B-Chat": "https://huggingface.co/SeaLLMs/SeaLLMs-v3-7B-Chat",
    "VIRNECT/llama-3-Korean-8B-V2": "https://huggingface.co/VIRNECT/llama-3-Korean-8B-V2",
    "MLP-KTLim/llama-3-Korean-Bllossom-8B": "https://huggingface.co/MLP-KTLim/llama-3-Korean-Bllossom-8B",
    "Magpie-Align/Llama-3-8B-Magpie-Align-v0.3": "https://huggingface.co/Magpie-Align/Llama-3-8B-Magpie-Align-v0.3",
    "cognitivecomputations/Llama-3-8B-Instruct-abliterated-v2": "https://huggingface.co/cognitivecomputations/Llama-3-8B-Instruct-abliterated-v2",
    "SkyOrbis/SKY-Ko-Llama3-8B-lora": "https://huggingface.co/SkyOrbis/SKY-Ko-Llama3-8B-lora",
    "4yo1/llama3-eng-ko-8b-sl5": "https://huggingface.co/4yo1/llama3-eng-ko-8b-sl5",
    "kimwooglae/WebSquareAI-Instruct-llama-3-8B-v0.5.39": "https://huggingface.co/kimwooglae/WebSquareAI-Instruct-llama-3-8B-v0.5.39",
    "ONS-AI-RESEARCH/ONS-SOLAR-10.7B-v1.2": "https://huggingface.co/ONS-AI-RESEARCH/ONS-SOLAR-10.7B-v1.2",
    "lcw99/llama-3-10b-it-kor-extented-chang-pro8": "https://huggingface.co/lcw99/llama-3-10b-it-kor-extented-chang-pro8",
    "BAAI/Infinity-Instruct-3M-0625-Yi-1.5-9B": "https://huggingface.co/BAAI/Infinity-Instruct-3M-0625-Yi-1.5-9B",
    "migtissera/Tess-2.0-Llama-3-8B": "https://huggingface.co/migtissera/Tess-2.0-Llama-3-8B",
    "BAAI/Infinity-Instruct-3M-0613-Mistral-7B": "https://huggingface.co/BAAI/Infinity-Instruct-3M-0613-Mistral-7B",
    "yeonwoo780/cydinfo-llama3-8b-lora-v01": "https://huggingface.co/yeonwoo780/cydinfo-llama3-8b-lora-v01",
    "vicgalle/ConfigurableSOLAR-10.7B": "https://huggingface.co/vicgalle/ConfigurableSOLAR-10.7B",
    "chihoonlee10/T3Q-ko-solar-jo-v1.0": "https://huggingface.co/chihoonlee10/T3Q-ko-solar-jo-v1.0",
    "Kukedlc/NeuralLLaMa-3-8b-ORPO-v0.4": "https://huggingface.co/Kukedlc/NeuralLLaMa-3-8b-ORPO-v0.4",
    "Edentns/DataVortexS-10.7B-dpo-v1.0": "https://huggingface.co/Edentns/DataVortexS-10.7B-dpo-v1.0",
    "SJ-Donald/SJ-SOLAR-10.7b-DPO": "https://huggingface.co/SJ-Donald/SJ-SOLAR-10.7b-DPO",
    "lemon-mint/gemma-ko-7b-it-v0.40": "https://huggingface.co/lemon-mint/gemma-ko-7b-it-v0.40",
    "GyuHyeonWkdWkdMan/naps-llama-3.1-8b-instruct-v0.3": "https://huggingface.co/GyuHyeonWkdWkdMan/naps-llama-3.1-8b-instruct-v0.3",
    "hyeogi/SOLAR-10.7B-v1.5": "https://huggingface.co/hyeogi/SOLAR-10.7B-v1.5",
    "etri-xainlp/llama3-8b-dpo_v1": "https://huggingface.co/etri-xainlp/llama3-8b-dpo_v1",
    "LDCC/LDCC-SOLAR-10.7B": "https://huggingface.co/LDCC/LDCC-SOLAR-10.7B",
    "chlee10/T3Q-Llama3-8B-Inst-sft1.0": "https://huggingface.co/chlee10/T3Q-Llama3-8B-Inst-sft1.0",
    "lemon-mint/gemma-ko-7b-it-v0.41": "https://huggingface.co/lemon-mint/gemma-ko-7b-it-v0.41",
    "chlee10/T3Q-Llama3-8B-sft1.0-dpo1.0": "https://huggingface.co/chlee10/T3Q-Llama3-8B-sft1.0-dpo1.0",
    "maywell/Synatra-7B-Instruct-v0.3-pre": "https://huggingface.co/maywell/Synatra-7B-Instruct-v0.3-pre",
    "UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter2": "https://huggingface.co/UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter2",
    "hwkwon/S-SOLAR-10.7B-v1.4": "https://huggingface.co/hwkwon/S-SOLAR-10.7B-v1.4",
    "12thD/ko-Llama-3-8B-sft-v0.3": "https://huggingface.co/12thD/ko-Llama-3-8B-sft-v0.3",
    "hkss/hk-SOLAR-10.7B-v1.4": "https://huggingface.co/hkss/hk-SOLAR-10.7B-v1.4",
    "lookuss/test-llilu": "https://huggingface.co/lookuss/test-llilu",
    "chihoonlee10/T3Q-ko-solar-dpo-v3.0": "https://huggingface.co/chihoonlee10/T3Q-ko-solar-dpo-v3.0",
    "chihoonlee10/T3Q-ko-solar-dpo-v1.0": "https://huggingface.co/chihoonlee10/T3Q-ko-solar-dpo-v1.0",
    "lcw99/llama-3-10b-wiki-240709-f": "https://huggingface.co/lcw99/llama-3-10b-wiki-240709-f",
    "Edentns/DataVortexS-10.7B-v0.4": "https://huggingface.co/Edentns/DataVortexS-10.7B-v0.4",
    "princeton-nlp/Llama-3-Instruct-8B-KTO": "https://huggingface.co/princeton-nlp/Llama-3-Instruct-8B-KTO",
    "spow12/kosolar_4.1_sft": "https://huggingface.co/spow12/kosolar_4.1_sft",
    "natong19/Qwen2-7B-Instruct-abliterated": "https://huggingface.co/natong19/Qwen2-7B-Instruct-abliterated",
    "megastudyedu/ME-dpo-7B-v1.1": "https://huggingface.co/megastudyedu/ME-dpo-7B-v1.1",
    "01-ai/Yi-1.5-9B-Chat-16K": "https://huggingface.co/01-ai/Yi-1.5-9B-Chat-16K",
    "Edentns/DataVortexS-10.7B-dpo-v0.1": "https://huggingface.co/Edentns/DataVortexS-10.7B-dpo-v0.1",
    "Alphacode-AI/AlphaMist7B-slr-v4-slow": "https://huggingface.co/Alphacode-AI/AlphaMist7B-slr-v4-slow",
    "chihoonlee10/T3Q-ko-solar-sft-dpo-v1.0": "https://huggingface.co/chihoonlee10/T3Q-ko-solar-sft-dpo-v1.0",
    "hwkwon/S-SOLAR-10.7B-v1.1": "https://huggingface.co/hwkwon/S-SOLAR-10.7B-v1.1",
    "DopeorNope/Dear_My_best_Friends-13B": "https://huggingface.co/DopeorNope/Dear_My_best_Friends-13B",
    "GyuHyeonWkdWkdMan/NAPS-llama-3.1-8b-instruct-v0.3.2": "https://huggingface.co/GyuHyeonWkdWkdMan/NAPS-llama-3.1-8b-instruct-v0.3.2",
    "PathFinderKR/Waktaverse-Llama-3-KO-8B-Instruct": "https://huggingface.co/PathFinderKR/Waktaverse-Llama-3-KO-8B-Instruct",
    "vicgalle/ConfigurableHermes-7B": "https://huggingface.co/vicgalle/ConfigurableHermes-7B",
    "maywell/PiVoT-10.7B-Mistral-v0.2": "https://huggingface.co/maywell/PiVoT-10.7B-Mistral-v0.2",
    "failspy/Meta-Llama-3-8B-Instruct-abliterated-v3": "https://huggingface.co/failspy/Meta-Llama-3-8B-Instruct-abliterated-v3",
    "lemon-mint/gemma-ko-7b-instruct-v0.50": "https://huggingface.co/lemon-mint/gemma-ko-7b-instruct-v0.50",
    "ENERGY-DRINK-LOVE/leaderboard_inst_v1.3_Open-Hermes_LDCC-SOLAR-10.7B_SFT": "https://huggingface.co/ENERGY-DRINK-LOVE/leaderboard_inst_v1.3_Open-Hermes_LDCC-SOLAR-10.7B_SFT",
    "maywell/PiVoT-0.1-early": "https://huggingface.co/maywell/PiVoT-0.1-early",
    "hwkwon/S-SOLAR-10.7B-v1.3": "https://huggingface.co/hwkwon/S-SOLAR-10.7B-v1.3",
    "werty1248/Llama-3-Ko-8B-Instruct-AOG": "https://huggingface.co/werty1248/Llama-3-Ko-8B-Instruct-AOG",
    "Alphacode-AI/AlphaMist7B-slr-v2": "https://huggingface.co/Alphacode-AI/AlphaMist7B-slr-v2",
    "maywell/koOpenChat-sft": "https://huggingface.co/maywell/koOpenChat-sft",
    "lemon-mint/gemma-7b-openhermes-v0.80": "https://huggingface.co/lemon-mint/gemma-7b-openhermes-v0.80",
    "VIRNECT/llama-3-Korean-8B-r-v1": "https://huggingface.co/VIRNECT/llama-3-Korean-8B-r-v1",
    "Alphacode-AI/AlphaMist7B-slr-v1": "https://huggingface.co/Alphacode-AI/AlphaMist7B-slr-v1",
    "Loyola/Mistral-7b-ITmodel": "https://huggingface.co/Loyola/Mistral-7b-ITmodel",
    "VIRNECT/llama-3-Korean-8B-r-v2": "https://huggingface.co/VIRNECT/llama-3-Korean-8B-r-v2",
    "NLPark/AnFeng_v3.1-Avocet": "https://huggingface.co/NLPark/AnFeng_v3.1-Avocet",
    "maywell/Synatra_TbST11B_EP01": "https://huggingface.co/maywell/Synatra_TbST11B_EP01",
    "GritLM/GritLM-7B-KTO": "https://huggingface.co/GritLM/GritLM-7B-KTO",
    "01-ai/Yi-34B-Chat": "https://huggingface.co/01-ai/Yi-34B-Chat",
    "ValiantLabs/Llama3.1-8B-ShiningValiant2": "https://huggingface.co/ValiantLabs/Llama3.1-8B-ShiningValiant2",
    "princeton-nlp/Llama-3-Base-8B-SFT-CPO": "https://huggingface.co/princeton-nlp/Llama-3-Base-8B-SFT-CPO",
    "hyokwan/hkcode_llama3_8b": "https://huggingface.co/hyokwan/hkcode_llama3_8b",
    "UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter3": "https://huggingface.co/UCLA-AGI/Llama-3-Instruct-8B-SPPO-Iter3",
    "yuntaeyang/SOLAR-10.7B-Instructlora_sftt-v1.0": "https://huggingface.co/yuntaeyang/SOLAR-10.7B-Instructlora_sftt-v1.0",
    "juungwon/Llama-3-cs-LoRA": "https://huggingface.co/juungwon/Llama-3-cs-LoRA",
    "gangyeolkim/llama-3-chat": "https://huggingface.co/gangyeolkim/llama-3-chat",
    "mncai/llama2-13b-dpo-v3": "https://huggingface.co/mncai/llama2-13b-dpo-v3",
    "maywell/Synatra-Zephyr-7B-v0.01": "https://huggingface.co/maywell/Synatra-Zephyr-7B-v0.01",
    "ENERGY-DRINK-LOVE/leaderboard_inst_v1.3_deup_LDCC-SOLAR-10.7B_SFT": "https://huggingface.co/ENERGY-DRINK-LOVE/leaderboard_inst_v1.3_deup_LDCC-SOLAR-10.7B_SFT",
    "juungwon/Llama-3-constructionsafety-LoRA": "https://huggingface.co/juungwon/Llama-3-constructionsafety-LoRA",
    "princeton-nlp/Mistral-7B-Base-SFT-SimPO": "https://huggingface.co/princeton-nlp/Mistral-7B-Base-SFT-SimPO",
    "moondriller/solar10B-eugeneparkthebestv2": "https://huggingface.co/moondriller/solar10B-eugeneparkthebestv2",
    "chlee10/T3Q-LLM3-Llama3-sft1.0-dpo1.0": "https://huggingface.co/chlee10/T3Q-LLM3-Llama3-sft1.0-dpo1.0",
    "Edentns/DataVortexS-10.7B-dpo-v1.7": "https://huggingface.co/Edentns/DataVortexS-10.7B-dpo-v1.7",
    "gamzadole/llama3_instruct_tuning_without_pretraing": "https://huggingface.co/gamzadole/llama3_instruct_tuning_without_pretraing",
    "saltlux/Ko-Llama3-Luxia-8B": "https://huggingface.co/saltlux/Ko-Llama3-Luxia-8B",
    "kimdeokgi/ko-pt-model-test1": "https://huggingface.co/kimdeokgi/ko-pt-model-test1",
    "maywell/Synatra-11B-Testbench-2": "https://huggingface.co/maywell/Synatra-11B-Testbench-2",
    "Danielbrdz/Barcenas-14b-Phi-3-medium-ORPO": "https://huggingface.co/Danielbrdz/Barcenas-14b-Phi-3-medium-ORPO",
    "vicgalle/Configurable-Mistral-7B": "https://huggingface.co/vicgalle/Configurable-Mistral-7B",
    "ENERGY-DRINK-LOVE/leaderboard_inst_v1.5_LDCC-SOLAR-10.7B_SFT": "https://huggingface.co/ENERGY-DRINK-LOVE/leaderboard_inst_v1.5_LDCC-SOLAR-10.7B_SFT",
    "beomi/Llama-3-Open-Ko-8B-Instruct-preview": "https://huggingface.co/beomi/Llama-3-Open-Ko-8B-Instruct-preview",
    "Edentns/DataVortexS-10.7B-dpo-v1.3": "https://huggingface.co/Edentns/DataVortexS-10.7B-dpo-v1.3",
    "spow12/Llama3_ko_4.2_sft": "https://huggingface.co/spow12/Llama3_ko_4.2_sft",
    "maywell/Llama-3-Ko-8B-Instruct": "https://huggingface.co/maywell/Llama-3-Ko-8B-Instruct",
    "T3Q-LLM/T3Q-LLM3-NC-v1.0": "https://huggingface.co/T3Q-LLM/T3Q-LLM3-NC-v1.0",
    "ehartford/dolphin-2.2.1-mistral-7b": "https://huggingface.co/ehartford/dolphin-2.2.1-mistral-7b",
    "hwkwon/S-SOLAR-10.7B-SFT-v1.3": "https://huggingface.co/hwkwon/S-SOLAR-10.7B-SFT-v1.3",
    "sel303/llama3-instruct-diverce-v2.0": "https://huggingface.co/sel303/llama3-instruct-diverce-v2.0",
    "4yo1/llama3-eng-ko-8b-sl3": "https://huggingface.co/4yo1/llama3-eng-ko-8b-sl3",
    "hkss/hk-SOLAR-10.7B-v1.1": "https://huggingface.co/hkss/hk-SOLAR-10.7B-v1.1",
    "Open-Orca/Mistral-7B-OpenOrca": "https://huggingface.co/Open-Orca/Mistral-7B-OpenOrca",
    "hyokwan/familidata": "https://huggingface.co/hyokwan/familidata",
    "uukuguy/zephyr-7b-alpha-dare-0.85": "https://huggingface.co/uukuguy/zephyr-7b-alpha-dare-0.85",
    "gwonny/nox-solar-10.7b-v4-kolon-all-5": "https://huggingface.co/gwonny/nox-solar-10.7b-v4-kolon-all-5",
    "shleeeee/mistral-ko-tech-science-v1": "https://huggingface.co/shleeeee/mistral-ko-tech-science-v1",
    "Deepnoid/deep-solar-eeve-KorSTS": "https://huggingface.co/Deepnoid/deep-solar-eeve-KorSTS",
    "AIdenU/Mistral-7B-v0.2-ko-Y24_v1.0": "https://huggingface.co/AIdenU/Mistral-7B-v0.2-ko-Y24_v1.0",
    "tlphams/gollm-tendency-45": "https://huggingface.co/tlphams/gollm-tendency-45",
    "realPCH/ko_solra_merge": "https://huggingface.co/realPCH/ko_solra_merge",
    "Cartinoe5930/original-KoRAE-13b": "https://huggingface.co/Cartinoe5930/original-KoRAE-13b",
    "GAI-LLM/Yi-Ko-6B-dpo-v5": "https://huggingface.co/GAI-LLM/Yi-Ko-6B-dpo-v5",
    "Minirecord/Mini_DPO_test02": "https://huggingface.co/Minirecord/Mini_DPO_test02",
    "AIJUUD/juud-Mistral-7B-dpo": "https://huggingface.co/AIJUUD/juud-Mistral-7B-dpo",
    "gwonny/nox-solar-10.7b-v4-kolon-all-10": "https://huggingface.co/gwonny/nox-solar-10.7b-v4-kolon-all-10",
    "jieunhan/TEST_MODEL": "https://huggingface.co/jieunhan/TEST_MODEL",
    "etri-xainlp/kor-llama2-13b-dpo": "https://huggingface.co/etri-xainlp/kor-llama2-13b-dpo",
    "ifuseok/yi-ko-playtus-instruct-v0.2": "https://huggingface.co/ifuseok/yi-ko-playtus-instruct-v0.2",
    "Cartinoe5930/original-KoRAE-13b-3ep": "https://huggingface.co/Cartinoe5930/original-KoRAE-13b-3ep",
    "Trofish/KULLM-RLHF": "https://huggingface.co/Trofish/KULLM-RLHF",
    "wkshin89/Yi-Ko-6B-Instruct-v1.0": "https://huggingface.co/wkshin89/Yi-Ko-6B-Instruct-v1.0",
    "momo/polyglot-ko-12.8b-Chat-QLoRA-Merge": "https://huggingface.co/momo/polyglot-ko-12.8b-Chat-QLoRA-Merge",
    "PracticeLLM/Custom-KoLLM-13B-v5": "https://huggingface.co/PracticeLLM/Custom-KoLLM-13B-v5",
    "BAAI/Infinity-Instruct-3M-0625-Yi-1.5-9B": "https://huggingface.co/BAAI/Infinity-Instruct-3M-0625-Yi-1.5-9B",
    "MRAIRR/minillama3_8b_all": "https://huggingface.co/MRAIRR/minillama3_8b_all",
    "failspy/Phi-3-medium-4k-instruct-abliterated-v3": "https://huggingface.co/failspy/Phi-3-medium-4k-instruct-abliterated-v3",
    "DILAB-HYU/koquality-polyglot-12.8b": "https://huggingface.co/DILAB-HYU/koquality-polyglot-12.8b",
    "kyujinpy/Korean-OpenOrca-v3": "https://huggingface.co/kyujinpy/Korean-OpenOrca-v3",
    "4yo1/llama3-eng-ko-8b": "https://huggingface.co/4yo1/llama3-eng-ko-8b",
    "4yo1/llama3-eng-ko-8": "https://huggingface.co/4yo1/llama3-eng-ko-8",
    "4yo1/llama3-eng-ko-8-llama": "https://huggingface.co/4yo1/llama3-eng-ko-8-llama",
    "PracticeLLM/Custom-KoLLM-13B-v2": "https://huggingface.co/PracticeLLM/Custom-KoLLM-13B-v2",
    "kyujinpy/KOR-Orca-Platypus-13B-v2": "https://huggingface.co/kyujinpy/KOR-Orca-Platypus-13B-v2",
    "ghost-x/ghost-7b-alpha": "https://huggingface.co/ghost-x/ghost-7b-alpha",
    "HumanF-MarkrAI/pub-llama-13B-v6": "https://huggingface.co/HumanF-MarkrAI/pub-llama-13B-v6",
    "nlpai-lab/kullm-polyglot-5.8b-v2": "https://huggingface.co/nlpai-lab/kullm-polyglot-5.8b-v2",
    "maywell/Synatra-42dot-1.3B": "https://huggingface.co/maywell/Synatra-42dot-1.3B",
    "yhkim9362/gemma-en-ko-7b-v0.1": "https://huggingface.co/yhkim9362/gemma-en-ko-7b-v0.1",
    "yhkim9362/gemma-en-ko-7b-v0.2": "https://huggingface.co/yhkim9362/gemma-en-ko-7b-v0.2",
    "daekeun-ml/Llama-2-ko-OpenOrca-gugugo-13B": "https://huggingface.co/daekeun-ml/Llama-2-ko-OpenOrca-gugugo-13B",
    "beomi/Yi-Ko-6B": "https://huggingface.co/beomi/Yi-Ko-6B",
    "jojo0217/ChatSKKU5.8B": "https://huggingface.co/jojo0217/ChatSKKU5.8B",
    "Deepnoid/deep-solar-v2.0.7": "https://huggingface.co/Deepnoid/deep-solar-v2.0.7",
    "01-ai/Yi-1.5-9B": "https://huggingface.co/01-ai/Yi-1.5-9B",
    "PracticeLLM/Custom-KoLLM-13B-v4": "https://huggingface.co/PracticeLLM/Custom-KoLLM-13B-v4",
    "nuebaek/komt_mistral_mss_user_0_max_steps_80": "https://huggingface.co/nuebaek/komt_mistral_mss_user_0_max_steps_80",
    "dltjdgh0928/lsh_finetune_v0.11": "https://huggingface.co/dltjdgh0928/lsh_finetune_v0.11",
    "shleeeee/mistral-7b-wiki": "https://huggingface.co/shleeeee/mistral-7b-wiki",
    "nayohan/polyglot-ko-5.8b-Inst": "https://huggingface.co/nayohan/polyglot-ko-5.8b-Inst",
    "ifuseok/sft-solar-10.7b-v1.1": "https://huggingface.co/ifuseok/sft-solar-10.7b-v1.1",
    "Junmai/KIT-5.8b": "https://huggingface.co/Junmai/KIT-5.8b",
    "heegyu/polyglot-ko-3.8b-chat": "https://huggingface.co/heegyu/polyglot-ko-3.8b-chat",
    "etri-xainlp/polyglot-ko-12.8b-instruct": "https://huggingface.co/etri-xainlp/polyglot-ko-12.8b-instruct",
    "OpenBuddy/openbuddy-mistral2-7b-v20.3-32k": "https://huggingface.co/OpenBuddy/openbuddy-mistral2-7b-v20.3-32k",
    "sh2orc/Llama-3-Korean-8B": "https://huggingface.co/sh2orc/Llama-3-Korean-8B",
    "Deepnoid/deep-solar-eeve-v2.0.0": "https://huggingface.co/Deepnoid/deep-solar-eeve-v2.0.0",
    "Herry443/Mistral-7B-KNUT-ref": "https://huggingface.co/Herry443/Mistral-7B-KNUT-ref",
    "heegyu/polyglot-ko-5.8b-chat": "https://huggingface.co/heegyu/polyglot-ko-5.8b-chat",
    "jungyuko/DAVinCI-42dot_LLM-PLM-1.3B-v1.5.3": "https://huggingface.co/jungyuko/DAVinCI-42dot_LLM-PLM-1.3B-v1.5.3",
    "DILAB-HYU/KoQuality-Polyglot-5.8b": "https://huggingface.co/DILAB-HYU/KoQuality-Polyglot-5.8b",
    "Byungchae/k2s3_test_0000": "https://huggingface.co/Byungchae/k2s3_test_0000",
    "migtissera/Tess-v2.5-Phi-3-medium-128k-14B": "https://huggingface.co/migtissera/Tess-v2.5-Phi-3-medium-128k-14B",
    "kyujinpy/Korean-OpenOrca-13B": "https://huggingface.co/kyujinpy/Korean-OpenOrca-13B",
    "kyujinpy/KO-Platypus2-13B": "https://huggingface.co/kyujinpy/KO-Platypus2-13B",
    "jin05102518/Astral-7B-Instruct-v0.01": "https://huggingface.co/jin05102518/Astral-7B-Instruct-v0.01",
    "Byungchae/k2s3_test_0002": "https://huggingface.co/Byungchae/k2s3_test_0002",
    "NousResearch/Nous-Hermes-llama-2-7b": "https://huggingface.co/NousResearch/Nous-Hermes-llama-2-7b",
    "kaist-ai/prometheus-13b-v1.0": "https://huggingface.co/kaist-ai/prometheus-13b-v1.0",
    "sel303/llama3-diverce-ver1.0": "https://huggingface.co/sel303/llama3-diverce-ver1.0",
    "NousResearch/Nous-Capybara-7B": "https://huggingface.co/NousResearch/Nous-Capybara-7B",
    "rrw-x2/KoSOLAR-10.7B-DPO-v1.0": "https://huggingface.co/rrw-x2/KoSOLAR-10.7B-DPO-v1.0",
    "Edentns/DataVortexS-10.7B-v0.2": "https://huggingface.co/Edentns/DataVortexS-10.7B-v0.2",
    "Jsoo/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6": "https://huggingface.co/Jsoo/Llama3-beomi-Open-Ko-8B-Instruct-preview-test6",
    "tlphams/gollm-instruct-all-in-one-v1": "https://huggingface.co/tlphams/gollm-instruct-all-in-one-v1",
    "Edentns/DataVortexTL-1.1B-v0.1": "https://huggingface.co/Edentns/DataVortexTL-1.1B-v0.1",
    "richard-park/llama3-pre1-ds": "https://huggingface.co/richard-park/llama3-pre1-ds",
    "ehartford/samantha-1.1-llama-33b": "https://huggingface.co/ehartford/samantha-1.1-llama-33b",
    "heegyu/LIMA-13b-hf": "https://huggingface.co/heegyu/LIMA-13b-hf",
    "heegyu/42dot_LLM-PLM-1.3B-mt": "https://huggingface.co/heegyu/42dot_LLM-PLM-1.3B-mt",
    "shleeeee/mistral-ko-7b-wiki-neft": "https://huggingface.co/shleeeee/mistral-ko-7b-wiki-neft",
    "EleutherAI/polyglot-ko-1.3b": "https://huggingface.co/EleutherAI/polyglot-ko-1.3b",
    "kyujinpy/Ko-PlatYi-6B-gu": "https://huggingface.co/kyujinpy/Ko-PlatYi-6B-gu",
    "sel303/llama3-diverce-ver1.6": "https://huggingface.co/sel303/llama3-diverce-ver1.6"    
}

def get_korea_models():
    """Korea 관련 모델 검색"""
    params = {
        "search": "korea",
        "full": "True",
        "config": "True",
        "limit": 1000
    }
    
    try:
        response = requests.get(
            "https://huggingface.co/api/models",
            headers={'Accept': 'application/json'},
            params=params
        )
        
        if response.status_code == 200:
            return response.json()
        else:
            print(f"Failed to fetch Korea models: {response.status_code}")
            return []
    except Exception as e:
        print(f"Error fetching Korea models: {str(e)}")
        return []

def get_all_models(limit=3000):
    """모든 모델과 Korea 관련 모델 가져오기"""
    all_models = []
    page_size = 1000  # API의 한 번 요청당 최대 크기
    
    # 여러 페이지에 걸쳐 데이터 수집
    for offset in range(0, limit, page_size):
        params = {
            'limit': min(page_size, limit - offset),
            'full': 'True',
            'config': 'True',
            'offset': offset
        }
        
        response = requests.get(
            "https://huggingface.co/api/models",
            headers={'Accept': 'application/json'},
            params=params
        )
        
        if response.status_code == 200:
            all_models.extend(response.json())
            print(f"Fetched models {offset+1} to {offset+len(response.json())}")
        else:
            print(f"Failed to fetch models at offset {offset}: {response.status_code}")
            break
    
    # Korea 검색 결과도 동일하게 확장
    korea_params = {
        "search": "korea",
        "full": "True",
        "config": "True",
        "limit": limit
    }
    
    korea_response = requests.get(
        "https://huggingface.co/api/models",
        headers={'Accept': 'application/json'},
        params=korea_params
    )
    
    if korea_response.status_code == 200:
        korea_models = korea_response.json()
        print(f"Fetched {len(korea_models)} Korea-related models")
        
        # 중복 제거하면서 Korea 모델 추가
        existing_ids = {model.get('id', '') for model in all_models}
        for model in korea_models:
            if model.get('id', '') not in existing_ids:
                all_models.append(model)
                existing_ids.add(model.get('id', ''))
    
    print(f"Total unique models: {len(all_models)}")
    return all_models[:limit]

def get_models_data(progress=gr.Progress()):
    def calculate_rank(model_id, all_global_models, korea_models):
        # 글로벌 순위 확인
        global_rank = next((idx for idx, m in enumerate(all_global_models, 1) 
                          if m.get('id', '').strip() == model_id.strip()), None)
        
        # Korea 모델인 경우
        is_korea = any(m.get('id', '').strip() == model_id.strip() for m in korea_models)
        
        if is_korea:
            # Korea 모델 중에서의 순위 확인
            korea_rank = next((idx for idx, m in enumerate(korea_models, 1)
                             if m.get('id', '').strip() == model_id.strip()), None)
            
            if korea_rank:
                return min(global_rank or 3001, korea_rank + 1000), True
        
        return global_rank if global_rank else 'Not in top 3000', is_korea

    try:
        progress(0, desc="Fetching models...")
        
        if not HF_TOKEN:
            fig = create_error_plot()
            error_html = """
            <div style='padding: 20px; background: #fee; border-radius: 10px; margin: 10px 0;'>
                <h3 style='color: #c00;'>⚠️ API 인증이 필요합니다</h3>
                <p>HuggingFace API 토큰이 설정되지 않았습니다. 완전한 기능을 사용하기 위해서는 API 토큰이 필요합니다.</p>
            </div>
            """
            empty_df = pd.DataFrame(columns=['Global Rank', 'Model ID', 'Title', 'Downloads', 'Likes', 'Korea Search', 'URL'])
            return fig, error_html, empty_df

        # 일반 모델과 Korea 관련 모델 모두 가져오기 (3000위까지)
        all_global_models = get_all_models(limit=3000)
        korea_models = get_korea_models()
        
        print(f"Total global models fetched: {len(all_global_models)}")
        print(f"Total Korea models fetched: {len(korea_models)}")
        
        # 모든 모델 통합 (중복 제거)
        all_models = all_global_models.copy()
        existing_ids = {model.get('id', '') for model in all_global_models}
        
        added_korea_models = 0
        for korea_model in korea_models:
            if korea_model.get('id', '') not in existing_ids:
                all_models.append(korea_model)
                existing_ids.add(korea_model.get('id', ''))
                added_korea_models += 1
        
        print(f"Added {added_korea_models} unique Korea models")
        print(f"Total combined models: {len(all_models)}")
        
        # 시각화를 위한 Figure 생성
        fig = go.Figure()
        
        # 순위 정보 수집
        filtered_models = []
        for model_id in target_models.keys():
            try:
                normalized_id = model_id.strip('/')
                model_url_api = f"https://huggingface.co/api/models/{normalized_id}"
                response = requests.get(
                    model_url_api, 
                    headers={'Accept': 'application/json'}
                )
                
                if response.status_code == 200:
                    model_data = response.json()
                    rank, is_korea = calculate_rank(model_id, all_global_models, korea_models)
                    
                    filtered_models.append({
                        'id': model_id,
                        'global_rank': rank,
                        'downloads': model_data.get('downloads', 0),
                        'likes': model_data.get('likes', 0),
                        'title': model_data.get('title', 'No Title'),
                        'is_korea': is_korea
                    })
                    
                    print(f"Model {model_id}: Rank={rank}, Is Korea={is_korea}")
                else:
                    filtered_models.append({
                        'id': model_id,
                        'global_rank': 'Not in top 3000',
                        'downloads': 0,
                        'likes': 0,
                        'title': 'No Title',
                        'is_korea': False
                    })
            except Exception as e:
                print(f"Error processing {model_id}: {str(e)}")
                continue

        # 순위로 정렬
        filtered_models.sort(key=lambda x: float('inf') if isinstance(x['global_rank'], str) else x['global_rank'])
        
        # 시각화 데이터 준비
        valid_models = [m for m in filtered_models if isinstance(m['global_rank'], (int, float))]
        
        if valid_models:
            ids = [m['id'] for m in valid_models]
            ranks = [m['global_rank'] for m in valid_models]
            
            fig.add_trace(go.Bar(
                x=ids,
                y=[3001 - r for r in ranks],  # Y축 범위 3000까지 확장
                text=[f"Rank: #{r}<br>Downloads: {format(m['downloads'], ',')}<br>Likes: {format(m['likes'], ',')}"
                      for r, m in zip(ranks, valid_models)],
                textposition='auto',
                marker_color=['rgba(255,0,0,0.6)' if m['is_korea'] else 'rgba(0,0,255,0.6)' 
                            for m in valid_models],
                opacity=0.8
            ))

            fig.update_layout(
                title="HuggingFace Models Global Rankings (Up to #3000)",
                xaxis_title="Model ID",
                yaxis_title="Global Rank",
                yaxis=dict(
                    ticktext=[f"#{i}" for i in range(1, 3001, 100)],
                    tickvals=[3001 - i for i in range(1, 3001, 100)],
                    range=[0, 3000]
                ),
                height=800,
                showlegend=False,
                template='plotly_white',
                xaxis_tickangle=-45
            )

        # HTML 카드 생성
        html_content = """
        <div style='padding: 20px; background: #f5f5f5;'>
            <h2 style='color: #2c3e50;'>Models Rankings (Up to #3000)</h2>
            <div style='display: grid; grid-template-columns: repeat(auto-fill, minmax(300px, 1fr)); gap: 20px;'>
        """
        
        for model in filtered_models:
            rank_display = f"Global Rank #{model['global_rank']}" if isinstance(model['global_rank'], (int, float)) else "Not in top 3000"
            korea_badge = "🇰🇷 Korea Search Result" if model['is_korea'] else ""
            
            html_content += f"""
            <div style='
                background: white;
                padding: 20px;
                border-radius: 10px;
                box-shadow: 0 2px 4px rgba(0,0,0,0.1);
                transition: transform 0.2s;
                {f"border: 2px solid #e74c3c;" if model['is_korea'] else ""}
            '>
                <h3 style='color: #34495e;'>{rank_display}</h3>
                <h4 style='color: #2c3e50;'>{model['id']}</h4>
                <p style='color: #e74c3c; font-weight: bold;'>{korea_badge}</p>
                <p style='color: #7f8c8d;'>⬇️ Downloads: {format(model['downloads'], ',')}</p>
                <p style='color: #7f8c8d;'>👍 Likes: {format(model['likes'], ',')}</p>
                <a href='{target_models[model['id']]}' 
                   target='_blank' 
                   style='
                    display: inline-block;
                    padding: 8px 16px;
                    background: #3498db;
                    color: white;
                    text-decoration: none;
                    border-radius: 5px;
                    transition: background 0.3s;
                   '>
                   Visit Model 🔗
                </a>
            </div>
            """
        
        html_content += "</div></div>"
        
        # 데이터프레임 생성
        df = pd.DataFrame([{
            'Global Rank': f"#{m['global_rank']}" if isinstance(m['global_rank'], (int, float)) else m['global_rank'],
            'Model ID': m['id'],
            'Title': m['title'],
            'Downloads': format(m['downloads'], ','),
            'Likes': format(m['likes'], ','),
            'Korea Search': '🇰🇷' if m['is_korea'] else '',
            'URL': target_models[m['id']]
        } for m in filtered_models])
        
        progress(1.0, desc="Complete!")
        return fig, html_content, df
        
    except Exception as e:
        print(f"Error in get_models_data: {str(e)}")
        error_fig = create_error_plot()
        error_html = f"""
        <div style='padding: 20px; background: #fee; border-radius: 10px; margin: 10px 0;'>
            <h3 style='color: #c00;'>⚠️ 오류가 발생했습니다</h3>
            <p>{str(e)}</p>
        </div>
        """
        empty_df = pd.DataFrame(columns=['Global Rank', 'Model ID', 'Title', 'Downloads', 'Likes', 'Korea Search', 'URL'])
        return error_fig, error_html, empty_df

        
# 관심 스페이스 URL 리스트와 정보
target_spaces = {

    "VIDraft/ChemGenesis": "https://huggingface.co/spaces/VIDraft/ChemGenesis", 
    "seawolf2357/ocrlatex": "https://huggingface.co/spaces/seawolf2357/ocrlatex",    
    "seawolf2357/img2vid": "https://huggingface.co/spaces/seawolf2357/img2vid",   
    "seawolf2357/sd-prompt-gen": "https://huggingface.co/spaces/seawolf2357/sd-prompt-gen",   

    "openfree/badassgi": "https://huggingface.co/spaces/openfree/badassgi",    
    "openfree/tarotcard": "https://huggingface.co/spaces/openfree/tarotcard",   
    "openfree/drqxab": "https://huggingface.co/spaces/openfree/drqxab",   
    "aiqcamp/Polaroid": "https://huggingface.co/spaces/aiqcamp/Polaroid",   
    "ginigen/cartoon": "https://huggingface.co/spaces/ginigen/cartoon",
    "ginigen/Book-Cover": "https://huggingface.co/spaces/ginigen/Book-Cover",
    "aiqcamp/fash": "https://huggingface.co/spaces/aiqcamp/fash",
    "gunship999/Korea-Daily-News": "https://huggingface.co/spaces/gunship999/Korea-Daily-News",
    "kolaslab/Quantum": "https://huggingface.co/spaces/kolaslab/Quantum",
    "openfree/webtoon": "https://huggingface.co/spaces/openfree/webtoon",
    "immunobiotech/ChicagoGallery": "https://huggingface.co/spaces/immunobiotech/ChicagoGallery",
    "immunobiotech/MetropolitanMuseum": "https://huggingface.co/spaces/immunobiotech/MetropolitanMuseum",    
    "immunobiotech/opensky": "https://huggingface.co/spaces/immunobiotech/opensky",
    "kolaslab/Audio-Visualizer": "https://huggingface.co/spaces/kolaslab/Audio-Visualizer",
    "kolaslab/Radio-Learning": "https://huggingface.co/spaces/kolaslab/Radio-Learning",
    "kolaslab/Future-Gallaxy": "https://huggingface.co/spaces/kolaslab/Future-Gallaxy",    
    "openfree/ProteinGenesis": "https://huggingface.co/spaces/openfree/ProteinGenesis",
    "openfree/2025saju": "https://huggingface.co/spaces/openfree/2025saju",
    "ginigen/Dokdo-membership": "https://huggingface.co/spaces/ginigen/Dokdo-membership",
    "VIDraft/eum": "https://huggingface.co/spaces/VIDraft/eum",
    "kolaslab/VisionART": "https://huggingface.co/spaces/kolaslab/VisionART",
    "aiqtech/FLUX-military": "https://huggingface.co/spaces/aiqtech/FLUX-military",
    "fantaxy/Rolls-Royce": "https://huggingface.co/spaces/fantaxy/Rolls-Royce",
    "seawolf2357/flux-korea-hanbok-lora": "https://huggingface.co/spaces/seawolf2357/flux-korea-hanbok-lora",
    "seawolf2357/flux-korea-palace-lora": "https://huggingface.co/spaces/seawolf2357/flux-korea-palace-lora",
    "aiqcamp/flux-cat-lora": "https://huggingface.co/spaces/aiqcamp/flux-cat-lora",
    "gunship999/SexyImages": "https://huggingface.co/spaces/gunship999/SexyImages",
    "aiqtech/flux-claude-monet-lora": "https://huggingface.co/spaces/aiqtech/flux-claude-monet-lora",
    "ginigen/CANVAS-o3": "https://huggingface.co/spaces/ginigen/CANVAS-o3",
    "kolaslab/world-sdr": "https://huggingface.co/spaces/kolaslab/world-sdr",
    "seawolf2357/3D-Avatar-Generator": "https://huggingface.co/spaces/seawolf2357/3D-Avatar-Generator",
    "fantaxy/playground25": "https://huggingface.co/spaces/fantaxy/playground25",
    "openfree/ultpixgen": "https://huggingface.co/spaces/openfree/ultpixgen",
    "kolaslab/VISION-NIGHT": "https://huggingface.co/spaces/kolaslab/VISION-NIGHT",
    "kolaslab/FLUX-WEB": "https://huggingface.co/spaces/kolaslab/FLUX-WEB",
    "seawolf2357/REALVISXL-V5": "https://huggingface.co/spaces/seawolf2357/REALVISXL-V5",
    "ginipick/Dokdo-multimodal": "https://huggingface.co/spaces/ginipick/Dokdo-multimodal",
    "ginigen/theater": "https://huggingface.co/spaces/ginigen/theater",
    "VIDraft/stock": "https://huggingface.co/spaces/VIDraft/stock",
    "fantos/flxcontrol": "https://huggingface.co/spaces/fantos/flxcontrol",
    "fantos/textcutobject": "https://huggingface.co/spaces/fantos/textcutobject",
    "ginipick/FLUX-Prompt-Generator": "https://huggingface.co/spaces/ginipick/FLUX-Prompt-Generator",
    "fantaxy/flxloraexp": "https://huggingface.co/spaces/fantaxy/flxloraexp",
    "fantos/flxloraexp": "https://huggingface.co/spaces/fantos/flxloraexp",
    "seawolf2357/flxloraexp": "https://huggingface.co/spaces/seawolf2357/flxloraexp",
    "ginipick/flxloraexp": "https://huggingface.co/spaces/ginipick/flxloraexp",
    "ginipick/FLUX-Prompt-Generator": "https://huggingface.co/spaces/ginipick/FLUX-Prompt-Generator",
    "ginigen/Dokdo": "https://huggingface.co/spaces/ginigen/Dokdo",
    "aiqcamp/imagemagic": "https://huggingface.co/spaces/aiqcamp/imagemagic",
    "openfree/ColorRevive": "https://huggingface.co/spaces/openfree/ColorRevive",
    "VIDraft/RAGOndevice": "https://huggingface.co/spaces/VIDraft/RAGOndevice",
    "gunship999/Radar-Bluetooth": "https://huggingface.co/spaces/gunship999/Radar-Bluetooth",
    "gunship999/WiFi-VISION": "https://huggingface.co/spaces/gunship999/WiFi-VISION",
    "gunship999/SONAR-Radar": "https://huggingface.co/spaces/gunship999/SONAR-Radar",
    "aiqcamp/AudioLlama": "https://huggingface.co/spaces/aiqcamp/AudioLlama",    
    "ginigen/FLUXllama-Multilingual": "https://huggingface.co/spaces/ginigen/FLUXllama-Multilingual",    
    "ginipick/ginimedi": "https://huggingface.co/spaces/ginipick/ginimedi",
    "ginipick/ginilaw": "https://huggingface.co/spaces/ginipick/ginilaw",
    "ginipick/ginipharm": "https://huggingface.co/spaces/ginipick/ginipharm",
    "ginipick/FitGen": "https://huggingface.co/spaces/ginipick/FitGen",
    "fantaxy/FLUX-Animations": "https://huggingface.co/spaces/fantaxy/FLUX-Animations",
    "fantaxy/Remove-Video-Background": "https://huggingface.co/spaces/fantaxy/Remove-Video-Background",
    "fantaxy/ofai-flx-logo": "https://huggingface.co/spaces/fantaxy/ofai-flx-logo",
    "fantaxy/flx-pulid": "https://huggingface.co/spaces/fantaxy/flx-pulid",
    "fantaxy/flx-upscale": "https://huggingface.co/spaces/fantaxy/flx-upscale",
    "aiqcamp/Fashion-FLUX": "https://huggingface.co/spaces/aiqcamp/Fashion-FLUX",
    "ginipick/StyleGen": "https://huggingface.co/spaces/ginipick/StyleGen",
    "openfree/StoryStar": "https://huggingface.co/spaces/openfree/StoryStar",
    "fantos/x-mas": "https://huggingface.co/spaces/fantos/x-mas",
    "openfree/Korean-Leaderboard": "https://huggingface.co/spaces/openfree/Korean-Leaderboard",
    "ginipick/FLUXllama": "https://huggingface.co/spaces/ginipick/FLUXllama",
    "ginipick/SORA-3D": "https://huggingface.co/spaces/ginipick/SORA-3D",
    "fantaxy/Sound-AI-SFX": "https://huggingface.co/spaces/fantaxy/Sound-AI-SFX",
    "fantos/flx8lora": "https://huggingface.co/spaces/fantos/flx8lora",
    "ginigen/Canvas": "https://huggingface.co/spaces/ginigen/Canvas",
    "fantaxy/erotica": "https://huggingface.co/spaces/fantaxy/erotica",
    "ginipick/time-machine": "https://huggingface.co/spaces/ginipick/time-machine",
    "aiqcamp/FLUX-VisionReply": "https://huggingface.co/spaces/aiqcamp/FLUX-VisionReply",
    "openfree/Tetris-Game": "https://huggingface.co/spaces/openfree/Tetris-Game",
    "openfree/everychat": "https://huggingface.co/spaces/openfree/everychat",
    "VIDraft/mouse1": "https://huggingface.co/spaces/VIDraft/mouse1",
    "kolaslab/alpha-go": "https://huggingface.co/spaces/kolaslab/alpha-go",
    "ginipick/text3d": "https://huggingface.co/spaces/ginipick/text3d",
    "openfree/trending-board": "https://huggingface.co/spaces/openfree/trending-board",
    "cutechicken/tankwar": "https://huggingface.co/spaces/cutechicken/tankwar",
    "openfree/game-jewel": "https://huggingface.co/spaces/openfree/game-jewel",
    "VIDraft/mouse-chat": "https://huggingface.co/spaces/VIDraft/mouse-chat",
    "ginipick/AccDiffusion": "https://huggingface.co/spaces/ginipick/AccDiffusion",
    "aiqtech/Particle-Accelerator-Simulation": "https://huggingface.co/spaces/aiqtech/Particle-Accelerator-Simulation",
    "openfree/GiniGEN": "https://huggingface.co/spaces/openfree/GiniGEN",
    "kolaslab/3DAudio-Spectrum-Analyzer": "https://huggingface.co/spaces/kolaslab/3DAudio-Spectrum-Analyzer",
    "openfree/trending-news-24": "https://huggingface.co/spaces/openfree/trending-news-24",
    "ginipick/Realtime-FLUX": "https://huggingface.co/spaces/ginipick/Realtime-FLUX",
    "VIDraft/prime-number": "https://huggingface.co/spaces/VIDraft/prime-number",
    "kolaslab/zombie-game": "https://huggingface.co/spaces/kolaslab/zombie-game",
    "fantos/miro-game": "https://huggingface.co/spaces/fantos/miro-game",
    "kolaslab/shooting": "https://huggingface.co/spaces/kolaslab/shooting",
    "VIDraft/Mouse-Hackathon": "https://huggingface.co/spaces/VIDraft/Mouse-Hackathon",
    "aiqmaster/stocksimulation": "https://huggingface.co/spaces/aiqmaster/stocksimulation",
    "aiqmaster/assetai": "https://huggingface.co/spaces/aiqmaster/assetai",
    "aiqmaster/stockai": "https://huggingface.co/spaces/aiqmaster/stockai",
    "cutechicken/TankWar3D": "https://huggingface.co/spaces/cutechicken/TankWar3D",
    "kolaslab/RC4-EnDecoder": "https://huggingface.co/spaces/kolaslab/RC4-EnDecoder",
    "kolaslab/simulator": "https://huggingface.co/spaces/kolaslab/simulator",
    "kolaslab/calculator": "https://huggingface.co/spaces/kolaslab/calculator",
    "aiqtech/kofaceid": "https://huggingface.co/spaces/aiqtech/kofaceid",
    "fantaxy/fastvideogena": "https://huggingface.co/spaces/fantaxy/fastvideogen",
    "fantos/cogvidx": "https://huggingface.co/spaces/fantos/cogvidx",
    "fantos/flxfashmodel": "https://huggingface.co/spaces/fantos/flxfashmodel",
    "fantos/kolcontrl": "https://huggingface.co/spaces/fantos/kolcontrl",
    "fantos/EveryText": "https://huggingface.co/spaces/fantos/EveryText",
    "aiqtech/cinevid": "https://huggingface.co/spaces/aiqtech/cinevid",
    "aiqtech/FLUX-Ghibli-Studio-LoRA": "https://huggingface.co/spaces/aiqtech/FLUX-Ghibli-Studio-LoRA",
    "aiqtech/flxgif": "https://huggingface.co/spaces/aiqtech/flxgif",
    "aiqtech/imaginpaint": "https://huggingface.co/spaces/aiqtech/imaginpaint",


    "upstage/open-ko-llm-leaderboard": "https://huggingface.co/spaces/upstage/open-ko-llm-leaderboard",
    "LGAI-EXAONE/EXAONE-3.5-Instruct-Demo": "https://huggingface.co/spaces/LGAI-EXAONE/EXAONE-3.5-Instruct-Demo",    
    "LeeSangHoon/HierSpeech_TTS": "https://huggingface.co/spaces/LeeSangHoon/HierSpeech_TTS",    
    "etri-vilab/Ko-LLaVA": "https://huggingface.co/spaces/etri-vilab/Ko-LLaVA",
    "etri-vilab/KOALA": "https://huggingface.co/spaces/etri-vilab/KOALA",
    "naver-clova-ix/donut-base-finetuned-cord-v2": "https://huggingface.co/spaces/naver-clova-ix/donut-base-finetuned-cord-v2",    
    "NCSOFT/VARCO_Arena": "https://huggingface.co/spaces/NCSOFT/VARCO_Arena"    
}

def get_spaces_data(sort_type="trending", progress=gr.Progress()):
    """스페이스 데이터 가져오기 (trending 또는 modes)"""
    url = "https://huggingface.co/api/spaces"
    params = {
        'full': 'true',
        'limit': 500
    }
    
    if sort_type == "modes":
        params['sort'] = 'likes'
    
    try:
        progress(0, desc=f"Fetching {sort_type} spaces data...")
        response = requests.get(url, params=params)
        response.raise_for_status()
        all_spaces = response.json()
        
        # 순위 정보 저장
        space_ranks = {}
        for idx, space in enumerate(all_spaces, 1):
            space_id = space.get('id', '')
            if space_id in target_spaces:
                space['rank'] = idx
                space_ranks[space_id] = space
        
        spaces = [space_ranks[space_id] for space_id in space_ranks.keys()]
        spaces.sort(key=lambda x: x['rank'])
        
        progress(0.3, desc="Creating visualization...")
        
        # 시각화 생성
        fig = go.Figure()
        
        # 데이터 준비
        ids = [space['id'] for space in spaces]
        ranks = [space['rank'] for space in spaces]
        likes = [space.get('likes', 0) for space in spaces]
        titles = [space.get('cardData', {}).get('title') or space.get('title', 'No Title') for space in spaces]
        
        # 막대 그래프 생성
        fig.add_trace(go.Bar(
            x=ids,
            y=ranks,
            text=[f"Rank: {r}<br>Title: {t}<br>Likes: {l}" 
                  for r, t, l in zip(ranks, titles, likes)],
            textposition='auto',
            marker_color='rgb(158,202,225)',
            opacity=0.8
        ))
        
        fig.update_layout(
            title={
                'text': f'Hugging Face Spaces {sort_type.title()} Rankings (Top 500)',
                'y':0.95,
                'x':0.5,
                'xanchor': 'center',
                'yanchor': 'top'
            },
            xaxis_title='Space ID',
            yaxis_title='Rank',
            yaxis=dict(
                autorange='reversed',  # Y축을 반전
                tickmode='array',
                ticktext=[str(i) for i in range(1, 501, 20)],  # 1부터 400까지 20 간격으로 표시
                tickvals=[i for i in range(1, 501, 20)],
                range=[1, 500]  # Y축 범위를 1부터 400까지로 설정
            ),
            height=800,
            showlegend=False,
            template='plotly_white',
            xaxis_tickangle=-45
        )
        
        progress(0.6, desc="Creating space cards...")
        
        # HTML 카드 생성
        html_content = f"""
        <div style='padding: 20px; background: #f5f5f5;'>
            <h2 style='color: #2c3e50;'>{sort_type.title()} Rankings</h2>
            <div style='display: grid; grid-template-columns: repeat(auto-fill, minmax(300px, 1fr)); gap: 20px;'>
        """
        
        for space in spaces:
            space_id = space['id']
            rank = space['rank']
            title = space.get('cardData', {}).get('title') or space.get('title', 'No Title')
            likes = space.get('likes', 0)

            html_content += f"""
            <div style='
                background: white;
                padding: 20px;
                border-radius: 10px;
                box-shadow: 0 2px 4px rgba(0,0,0,0.1);
                transition: transform 0.2s;
            '>
                <h3 style='color: #34495e;'>Rank #{rank} - {space_id}</h3>
                <h4 style='
                    color: #2980b9;
                    margin: 10px 0;
                    font-size: 1.2em;
                    font-weight: bold;
                    text-shadow: 1px 1px 2px rgba(0,0,0,0.1);
                    background: linear-gradient(to right, #3498db, #2980b9);
                    -webkit-background-clip: text;
                    -webkit-text-fill-color: transparent;
                    padding: 5px 0;
                '>{title}</h4>
                <p style='color: #7f8c8d; margin-bottom: 10px;'>👍 Likes: {likes}</p>
                <a href='{target_spaces[space_id]}' 
                   target='_blank' 
                   style='
                    display: inline-block;
                    padding: 8px 16px;
                    background: #3498db;
                    color: white;
                    text-decoration: none;
                    border-radius: 5px;
                    transition: background 0.3s;
                   '>
                   Visit Space 🔗
                </a>
            </div>
            """
        
        html_content += "</div></div>"
        
        # 데이터프레임 생성
        df = pd.DataFrame([{
            'Rank': space['rank'],
            'Space ID': space['id'],
            'Title': space.get('cardData', {}).get('title') or space.get('title', 'No Title'),
            'Likes': space.get('likes', 0),
            'URL': target_spaces[space['id']]
        } for space in spaces])
        
        progress(1.0, desc="Complete!")
        return fig, html_content, df
        
    except Exception as e:
        print(f"Error in get_spaces_data: {str(e)}")
        error_html = f'<div style="color: red; padding: 20px;">Error: {str(e)}</div>'
        error_plot = create_error_plot()
        return error_plot, error_html, pd.DataFrame()


def create_trend_visualization(spaces_data):
    if not spaces_data:
        return create_error_plot()
    
    fig = go.Figure()
    
    # 순위 데이터 준비
    ranks = []
    for idx, space in enumerate(spaces_data, 1):
        space_id = space.get('id', '')
        if space_id in target_spaces:
            ranks.append({
                'id': space_id,
                'rank': idx,
                'likes': space.get('likes', 0),
                'title': space.get('title', 'N/A'),
                'views': space.get('views', 0)
            })
    
    if not ranks:
        return create_error_plot()
    
    # 순위별로 정렬
    ranks.sort(key=lambda x: x['rank'])
    
    # 플롯 데이터 생성
    ids = [r['id'] for r in ranks]
    rank_values = [r['rank'] for r in ranks]
    likes = [r['likes'] for r in ranks]
    views = [r['views'] for r in ranks]
    
    # 막대 그래프 생성
    fig.add_trace(go.Bar(
        x=ids,
        y=rank_values,
        text=[f"Rank: {r}<br>Likes: {l}<br>Views: {v}" for r, l, v in zip(rank_values, likes, views)],
        textposition='auto',
        marker_color='rgb(158,202,225)',
        opacity=0.8
    ))
    
    fig.update_layout(
        title={
            'text': 'Current Trending Ranks (All Target Spaces)',
            'y':0.95,
            'x':0.5,
            'xanchor': 'center',
            'yanchor': 'top'
        },
        xaxis_title='Space ID',
        yaxis_title='Trending Rank',
        yaxis_autorange='reversed',
        height=800,
        showlegend=False,
        template='plotly_white',
        xaxis_tickangle=-45
    )
    
    return fig

# 토큰이 없는 경우를 위한 대체 함수
def get_trending_spaces_without_token():
    try:
        url = "https://huggingface.co/api/spaces"
        params = {
            'sort': 'likes',
            'direction': -1,
            'limit': 500,
            'full': 'true'
        }
        
        response = requests.get(url, params=params)
        
        if response.status_code == 200:
            return response.json()
        else:
            print(f"API 요청 실패 (토큰 없음): {response.status_code}")
            print(f"Response: {response.text}")
            return None
    except Exception as e:
        print(f"API 호출 중 에러 발생 (토큰 없음): {str(e)}")
        return None

# API 토큰 설정 및 함수 선택
if not HF_TOKEN:
    get_trending_spaces = get_trending_spaces_without_token



def create_error_plot():
    fig = go.Figure()
    fig.add_annotation(
        text="데이터를 불러올 수 없습니다.\n(API 인증이 필요합니다)",
        xref="paper",
        yref="paper",
        x=0.5,
        y=0.5,
        showarrow=False,
        font=dict(size=20)
    )
    fig.update_layout(
        title="Error Loading Data",
        height=400
    )
    return fig


def create_space_info_html(spaces_data):
    if not spaces_data:
        return "<div style='padding: 20px;'><h2>데이터를 불러오는데 실패했습니다.</h2></div>"
    
    html_content = """
    <div style='padding: 20px;'>
    <h2 style='color: #2c3e50;'>Current Trending Rankings</h2>
    <div style='display: grid; grid-template-columns: repeat(auto-fill, minmax(300px, 1fr)); gap: 20px;'>
    """
    
    # 모든 target spaces를 포함하도록 수정
    for space_id in target_spaces.keys():
        space_info = next((s for s in spaces_data if s.get('id') == space_id), None)
        if space_info:
            rank = next((idx for idx, s in enumerate(spaces_data, 1) if s.get('id') == space_id), 'N/A')
            html_content += f"""
            <div style='
                background: white;
                padding: 20px;
                border-radius: 10px;
                box-shadow: 0 2px 4px rgba(0,0,0,0.1);
                transition: transform 0.2s;
            '>
                <h3 style='color: #34495e;'>#{rank} - {space_id}</h3>
                <p style='color: #7f8c8d;'>👍 Likes: {space_info.get('likes', 'N/A')}</p>
                <p style='color: #7f8c8d;'>👀 Views: {space_info.get('views', 'N/A')}</p>
                <p style='color: #2c3e50;'>{space_info.get('title', 'N/A')}</p>
                <p style='color: #7f8c8d; font-size: 0.9em;'>{space_info.get('description', 'N/A')[:100]}...</p>
                <a href='{target_spaces[space_id]}' 
                   target='_blank' 
                   style='
                    display: inline-block;
                    padding: 8px 16px;
                    background: #3498db;
                    color: white;
                    text-decoration: none;
                    border-radius: 5px;
                    transition: background 0.3s;
                   '>
                   Visit Space 🔗
                </a>
            </div>
            """
        else:
            html_content += f"""
            <div style='
                background: #f8f9fa;
                padding: 20px;
                border-radius: 10px;
                box-shadow: 0 2px 4px rgba(0,0,0,0.1);
            '>
                <h3 style='color: #34495e;'>{space_id}</h3>
                <p style='color: #7f8c8d;'>Not in trending</p>
                <a href='{target_spaces[space_id]}' 
                   target='_blank' 
                   style='
                    display: inline-block;
                    padding: 8px 16px;
                    background: #95a5a6;
                    color: white;
                    text-decoration: none;
                    border-radius: 5px;
                   '>
                   Visit Space 🔗
                </a>
            </div>
            """
    
    html_content += "</div></div>"
    return html_content

def create_data_table(spaces_data):
    if not spaces_data:
        return pd.DataFrame()
    
    rows = []
    for idx, space in enumerate(spaces_data, 1):
        space_id = space.get('id', '')
        if space_id in target_spaces:
            rows.append({
                'Rank': idx,
                'Space ID': space_id,
                'Likes': space.get('likes', 'N/A'),
                'Title': space.get('title', 'N/A'),
                'URL': target_spaces[space_id]
            })
    
    return pd.DataFrame(rows)

def refresh_data():
    spaces_data = get_trending_spaces()
    if spaces_data:
        plot = create_trend_visualization(spaces_data)
        info = create_space_info_html(spaces_data)
        df = create_data_table(spaces_data)
        return plot, info, df
    else:
        return create_error_plot(), "<div>API 인증이 필요합니다.</div>", pd.DataFrame()



def create_registration_bar_chart(data, type_name="Spaces"):
    try:
        # TOP 기준 설정
        top_limit = 500 if type_name == "Spaces" else 3000
        
        # DataFrame인 경우 처리
        if isinstance(data, pd.DataFrame):
            if type_name == "Models":
                # 3000위 이내의 모델만 필터링
                data = data[data['Global Rank'].apply(lambda x: isinstance(x, (int, float)) or (isinstance(x, str) and x.startswith('#')))]
                data = data[data['Global Rank'].apply(lambda x: int(str(x).replace('#', '')) if isinstance(x, str) else x) <= top_limit]
            elif type_name == "Spaces":
                # 500위 이내의 스페이스만 필터링
                data = data[data['Rank'].apply(lambda x: isinstance(x, (int, float))) & (data['Rank'] <= top_limit)]
            
            # ID 컬럼 선택
            id_column = 'Space ID' if type_name == "Spaces" else 'Model ID'
            registrations = data[id_column].apply(lambda x: x.split('/')[0]).value_counts()
        else:
            # 리스트나 다른 형태의 데이터인 경우 처리
            registrations = {}
            for item in data:
                if isinstance(item, dict):
                    rank = item.get('global_rank' if type_name == "Models" else 'rank')
                    if isinstance(rank, str) or rank > top_limit:
                        continue
                    creator = item.get('id', '').split('/')[0]
                    registrations[creator] = registrations.get(creator, 0) + 1
            registrations = pd.Series(registrations)
        
        # 정렬된 데이터 준비
        registrations = registrations.sort_values(ascending=False)
        
        fig = go.Figure(data=[go.Bar(
            x=registrations.index,
            y=registrations.values,
            text=registrations.values,
            textposition='auto',
            marker_color='#FF6B6B'
        )])
        
        fig.update_layout(
            title=f"Korean {type_name} Registrations by Creator (Top {top_limit})",
            xaxis_title="Creator ID",
            yaxis_title="Number of Registrations",
            showlegend=False,
            height=400,
            width=700
        )
        
        return fig
    except Exception as e:
        print(f"Error in create_registration_bar_chart: {str(e)}")
        return go.Figure()

def create_pie_chart(data, total_count, type_name="Spaces"):
    try:
        # TOP 기준 설정
        top_limit = 500 if type_name == "Spaces" else 3000
        
        # DataFrame인 경우 처리
        if isinstance(data, pd.DataFrame):
            if type_name == "Models":
                # 3000위 이내의 모델만 필터링
                data = data[data['Global Rank'].apply(lambda x: isinstance(x, (int, float)) or (isinstance(x, str) and x.startswith('#')))]
                data = data[data['Global Rank'].apply(lambda x: int(str(x).replace('#', '')) if isinstance(x, str) else x) <= top_limit]
            elif type_name == "Spaces":
                # 500위 이내의 스페이스만 필터링
                data = data[data['Rank'].apply(lambda x: isinstance(x, (int, float))) & (data['Rank'] <= top_limit)]
            korean_count = len(data)
        else:
            # 리스트나 다른 형태의 데이터인 경우 처리
            if type_name == "Models":
                korean_count = sum(1 for item in data if isinstance(item.get('global_rank'), (int, float)) and item.get('global_rank') <= top_limit)
            else:
                korean_count = sum(1 for item in data if isinstance(item.get('rank'), (int, float)) and item.get('rank') <= top_limit)
        
        other_count = total_count - korean_count
        
        fig = go.Figure(data=[go.Pie(
            labels=[f'Korean {type_name} in Top {top_limit}', f'Other {type_name} in Top {top_limit}'],
            values=[korean_count, other_count],
            hole=.3,
            marker_colors=['#FF6B6B', '#4ECDC4'],
            textinfo='percent+value',
            hovertemplate="<b>%{label}</b><br>" +
                         "Count: %{value}<br>" +
                         "Percentage: %{percent}<br>"
        )])
        
        fig.update_layout(
            title=f"Korean vs Other {type_name} Distribution (Top {top_limit})",
            showlegend=True,
            height=400,
            width=500
        )
        
        return fig
    except Exception as e:
        print(f"Error in create_pie_chart: {str(e)}")
        return go.Figure()

def refresh_all_data():
    spaces_results = get_spaces_data("trending")
    models_results = get_models_data()
    
    # Spaces 차트 생성
    spaces_pie = create_pie_chart(spaces_results[2], 500, "Spaces")
    spaces_bar = create_registration_bar_chart(spaces_results[2], "Spaces")
    
    # Models 차트 생성
    models_pie = create_pie_chart(models_results[2], 3000, "Models")
    models_bar = create_registration_bar_chart(models_results[2], "Models")
    
    return [
        spaces_results[0], spaces_results[1], spaces_results[2],
        spaces_pie, spaces_bar,
        models_results[0], models_results[1], models_results[2],
        models_pie, models_bar
    ]


with gr.Blocks(theme="Yntec/HaleyCH_Theme_Orange", css="""
    #spaces_pie, #models_pie {
        min-height: 400px;
        border-radius: 10px;
        box-shadow: 0 2px 4px rgba(0,0,0,0.1);
    }
    #spaces_bar, #models_bar {
        min-height: 400px;
        border-radius: 10px;
        box-shadow: 0 2px 4px rgba(0,0,0,0.1);
    }
""") as demo:

    gr.Markdown("""
    # 🤗 허깅페이스 '한국(언어) 리더보드'
    HuggingFace가 제공하는 Spaces와 Models 실시간 인기 순위 반영하여 '한국인(기업/언어)'의 리스트(공개,검색,리더보드 등)만 분석. (c)'한국인공지능진흥협회' / 요청: [email protected]
    """)
    
    # 이미지와 설명 추가
    gr.Markdown("""
    ### [Hot NEWS] 허깅페이스 선정 12월 'TOP 12'에 한국 'ginipick'의 'FLUXllama'와 'Text3D' 2종이 선정됨
    """)
    gr.Image("HF-TOP12.png", show_label=False)
    
    # 새로 고침 버튼 (기존 코드)
    refresh_btn = gr.Button("🔄 새로 고침", variant="primary")
    

    with gr.Tab("Spaces Trending"):
        trending_plot = gr.Plot()
        with gr.Row():
            # 원형 그래프와 막대 그래프를 위한 컨테이너 추가
            with gr.Column(scale=1):
                spaces_pie_chart = gr.Plot(
                    label="Korean Spaces Distribution",
                    elem_id="spaces_pie"
                )
            with gr.Column(scale=2):
                spaces_bar_chart = gr.Plot(
                    label="Registrations by Creator",
                    elem_id="spaces_bar"
                )
        trending_info = gr.HTML()
        trending_df = gr.DataFrame(
            headers=["Rank", "Space ID", "Title", "Likes", "URL"],
            datatype=["number", "str", "str", "number", "str"],
            row_count=(10, "dynamic")
        )
    
    with gr.Tab("Models Trending"):
        models_plot = gr.Plot()
        with gr.Row():
            # 원형 그래프와 막대 그래프를 위한 컨테이너 추가
            with gr.Column(scale=1):
                models_pie_chart = gr.Plot(
                    label="Korean Models Distribution",
                    elem_id="models_pie"
                )
            with gr.Column(scale=2):
                models_bar_chart = gr.Plot(
                    label="Registrations by Creator",
                    elem_id="models_bar"
                )
        models_info = gr.HTML()
        models_df = gr.DataFrame(
            headers=["Global Rank", "Model ID", "Title", "Downloads", "Likes", "Korea Search", "URL"],
            datatype=["str", "str", "str", "str", "str", "str", "str"],
            row_count=(10, "dynamic")
        )

    def refresh_all_data():
        try:
            spaces_results = get_spaces_data("trending")
            models_results = get_models_data()
            
            # Spaces 차트 생성
            spaces_pie = create_pie_chart(spaces_results[2], 500, "Spaces")
            spaces_bar = create_registration_bar_chart(spaces_results[2], "Spaces")
            
            # Models 차트 생성
            models_pie = create_pie_chart(models_results[2], 3000, "Models")
            models_bar = create_registration_bar_chart(models_results[2], "Models")
            
            return [
                spaces_results[0], spaces_results[1], spaces_results[2],
                spaces_pie, spaces_bar,
                models_results[0], models_results[1], models_results[2],
                models_pie, models_bar
            ]
        except Exception as e:
            print(f"Error in refresh_all_data: {str(e)}")
            # 에러 발생 시 기본값 반환
            return [None] * 10

    # 새로고침 버튼 클릭 이벤트 핸들러
    refresh_btn.click(
        fn=refresh_all_data,
        outputs=[
            trending_plot, trending_info, trending_df,
            spaces_pie_chart, spaces_bar_chart,
            models_plot, models_info, models_df,
            models_pie_chart, models_bar_chart
        ]
    )
    
    # 초기 데이터 로드
    try:
        initial_data = refresh_all_data()
        
        # 초기값 설정
        trending_plot.value = initial_data[0]
        trending_info.value = initial_data[1]
        trending_df.value = initial_data[2]
        spaces_pie_chart.value = initial_data[3]
        spaces_bar_chart.value = initial_data[4]
        models_plot.value = initial_data[5]
        models_info.value = initial_data[6]
        models_df.value = initial_data[7]
        models_pie_chart.value = initial_data[8]
        models_bar_chart.value = initial_data[9]
    except Exception as e:
        print(f"Error loading initial data: {str(e)}")
        gr.Warning("초기 데이터 로드 중 오류가 발생했습니다.")

# Gradio 앱 실행
demo.launch(
    server_name="0.0.0.0",
    server_port=7860,
    share=False,
    show_error=True
)