text
stringlengths
0
473k
[SOURCE: https://www.lomdimbareshet.net/%d7%9e%d7%aa%d7%9e%d7%98%d7%99%d7%a7%d7%94-%d7%9b%d7%99%d7%aa%d7%94-%d7%97/] | [TOKENS: 913]
ืžืชืžื˜ื™ืงื” ืœื›ื™ืชื” ื—' ืชื•ื›ื ื™ืช ื”ืœื™ืžื•ื“ื™ื ื‘ืžืชืžื˜ื™ืงื” ืœื›ื™ืชื” ื—' ืžื•ืจื›ื‘ืช ืžืฉืœื•ืฉื” ืชื—ื•ืžื™ื: ืืœื’ื‘ืจื™ , ื’ืื•ืžื˜ืจื™ ื•ืžืกืคืจื™ ( ื—ื™ืฉื•ื‘ื™ ).ืœืคื ื™ื›ื ื ื•ืฉืื™ ื”ืชื•ื›ื ื™ืช ื‘ืชื•ืกืคืช ืกืจื˜ื•ื ื™ ื”ืกื‘ืจ ืœื—ื•ืžืจ ื”ืœื™ืžื•ื“. ื‘ืชื—ืชื™ืช ื”ื“ืฃ ืชืžืฆืื• ืžื‘ื—ื ื™ ืžื™ืฆ"ื‘ ืžืฉื ื™ื ืงื•ื“ืžื•ืช. ืœืžื™ื“ื” ืžื”ื ื”. ืคื•ื ืงืฆื™ื” ืงื•ื•ื™ืช ืื™ ืฉื™ื•ื•ื™ื ื•ืช ืงื•ื•ื™ื™ื ืžืฉื•ื•ืื•ืช ืขื ื ืขืœื ืื—ื“ ืฉืืœื•ืช ืžื™ืœื•ืœื™ื•ืช ืขื ื ืขืœื ืื—ื“ ื˜ื›ื ื™ืงื” ืืœื’ื‘ืจื™ืช ืžืขืจื›ืช ืžืฉื•ื•ืื•ืช ืขื ืฉื ื™ ื ืขืœืžื™ื ืฉืืœื•ืช ืžื™ืœื•ืœื™ื•ืช ืขื ืฉื ื™ ื ืขืœืžื™ื ืขืจืš ืžื•ื—ืœื˜ ืื™ ืฉื™ื•ื•ื™ื•ื ื•ืช ื™ื—ืก ื‘ื™ืŸ ืžืกืคืจื™ื ื•ื™ื—ืก ื™ืฉื™ืจ ืคื•ืจืคื•ืจืฆื™ื” ืงื ื” ืžื™ื“ื” ื™ื—ืก ื”ืคื•ืš ืื—ื•ื–ื™ื ืกื˜ื˜ื™ืกื˜ื™ืงื” ืชื™ืื•ืจื™ืช ื”ืกืชื‘ืจื•ืช ืฉื•ืจืฉ ืจื™ื‘ื•ืขื™ ื•ืžืกืคืจ ืื™ ืจืฆื™ื•ื ืืœื™ ืžืฉื•ืœืฉื™ื ื—ื•ืคืคื™ื ืชื™ื›ื•ืŸ ื‘ืžืฉื•ืœืฉ ืžืฉื•ืœืฉ ืฉื•ื•ื” ืฉื•ืงื™ื™ื ืžืฉื•ืœืฉื™ื ื“ื•ืžื™ื ื•ืžืฆื•ืœืขื™ื ื“ื•ืžื™ื ืžืฉื•ืœืฉ ื™ืฉืจ ื–ื•ื•ื™ืช ืžืฉืคื˜ ืคื™ืชื’ื•ืจืก ื‘ืžื™ืฉื•ืจ ื•ื‘ืžืจื—ื‘ ื’ืœื™ืœ ืžื‘ื—ื ื™ ืžื™ืฆ"ื‘ ื‘ืžืชืžื˜ื™ืงื” ืœืชืœืžื™ื“ื™ ื›ื™ืชื•ืช ื—' ื›ื•ืœืœื™ื ื—ื•ืžืจ ืžืชื•ื›ื ื™ืช ื”ืœื™ืžื•ื“ื™ื ืœื›ื™ืชื” ื–' ื•ืžืฉื ื™ ื”ืฉืœื™ืฉื™ื ื”ืจืืฉื•ื ื™ื ืฉืœ ื›ื™ืชื” ื—'.ืžื‘ื—ืŸ ื”ืžื™ืฆ"ื‘ ื”ื•ื ืื—ื™ื“ ืœื›ืœืœ ื”ืชืœืžื™ื“ื™ื. ืœื”ืœืŸ ืžื‘ื—ื ื™ ืžื™ืฆ"ื‘ ืžื”ืฉื ื™ื ื”ืื—ืจื•ื ื•ืช ื•ืคืชืจื•ืŸ ื‘ื•ื™ื“ืื• ืœืžื‘ื—ืŸ ืžืฉื ืช 2012. ืฆืจื• ืงืฉืจ - ื ืฉืžื— ืœืขื–ื•ืจ. ยฉ 2024 ื›ืœ ื”ื–ื›ื•ื™ื•ืช ืฉืžื•ืจื•ืช. ืœื•ืžื“ื™ื ื‘ืจืฉืช - ื“ื•ื“ื• ื’ื•ืœื“ืฉื˜ื™ื™ืŸ ืขืจื•ืฅ "ืœื•ืžื“ื™ื ื‘ืจืฉืช" ืœื™ืžื•ื“ ืขืฆืžื™
========================================
[SOURCE: https://huggingface.co/models?library=safetensors] | [TOKENS: 918]
Models Qwen/Qwen3.5-397B-A17B Image-Text-to-Text โ€ข Updated 1 day ago โ€ข 133k โ€ข โ€ข 802 MiniMaxAI/MiniMax-M2.5 Text Generation โ€ข Updated 5 days ago โ€ข 173k โ€ข โ€ข 816 nvidia/personaplex-7b-v1 Audio-to-Audio โ€ข Updated 6 days ago โ€ข 539k โ€ข 2.1k zai-org/GLM-5 Text Generation โ€ข 754B โ€ข Updated 8 days ago โ€ข 177k โ€ข โ€ข 1.39k Nanbeige/Nanbeige4.1-3B Text Generation โ€ข 4B โ€ข Updated 2 days ago โ€ข 130k โ€ข โ€ข 636 FireRedTeam/FireRed-Image-Edit-1.0 Image-to-Image โ€ข Updated 7 days ago โ€ข 2.15k โ€ข โ€ข 228 Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice Text-to-Speech โ€ข Updated 23 days ago โ€ข 933k โ€ข 1.12k nineninesix/kani-tts-2-en Text-to-Speech โ€ข 0.4B โ€ข Updated 2 days ago โ€ข 2.59k โ€ข 163 xgen-universe/Capybara Any-to-Any โ€ข Updated 1 day ago โ€ข 141 OpenMOSS-Team/MOSS-TTS Text-to-Speech โ€ข 8B โ€ข Updated 8 days ago โ€ข 41.2k โ€ข 289 moonshotai/Kimi-K2.5 Image-Text-to-Text โ€ข Updated 16 days ago โ€ข 1.07M โ€ข โ€ข 2.06k jdopensource/JoyAI-LLM-Flash Text Generation โ€ข 49B โ€ข Updated 3 days ago โ€ข 807 โ€ข 128 Qwen/Qwen3-Coder-Next Text Generation โ€ข Updated 18 days ago โ€ข 415k โ€ข โ€ข 933 nvidia/NVIDIA-Nemotron-Nano-9B-v2-Japanese Text Generation โ€ข 9B โ€ข Updated 1 day ago โ€ข 5.14k โ€ข 92 Fortytwo-Network/Strand-Rust-Coder-14B-v1 Text Generation โ€ข 15B โ€ข Updated Jan 5 โ€ข 627 โ€ข 103 mistralai/Voxtral-Mini-4B-Realtime-2602 Automatic Speech Recognition โ€ข Updated 3 days ago โ€ข 101k โ€ข 596 Zyphra/ZUNA 0.4B โ€ข Updated 3 days ago โ€ข 747 โ€ข 84 inclusionAI/Ling-2.5-1T Text Generation โ€ข Updated 5 days ago โ€ข 1.64k โ€ข 80 zai-org/GLM-OCR Image-to-Text โ€ข Updated 12 days ago โ€ข 1.32M โ€ข 1.1k inclusionAI/Ring-2.5-1T Text Generation โ€ข Updated 6 days ago โ€ข 5.35k โ€ข 215 shallowdream204/BitDance-14B-16x Text-to-Image โ€ข 15B โ€ข Updated 3 days ago โ€ข 117 โ€ข 59 AIDC-AI/Ovis2.6-30B-A3B Image-Text-to-Text โ€ข 31B โ€ข Updated 9 days ago โ€ข 11.1k โ€ข 132 inclusionAI/Ming-flash-omni-2.0 Any-to-Any โ€ข Updated 9 days ago โ€ข 7.86k โ€ข 241 CohereLabs/tiny-aya-global Text Generation โ€ข 3B โ€ข Updated 2 days ago โ€ข 988 โ€ข 47 openbmb/MiniCPM-SALA Text Generation โ€ข Updated 10 days ago โ€ข 4.66k โ€ข 470 Qwen/Qwen3.5-397B-A17B-FP8 Image-Text-to-Text โ€ข 403B โ€ข Updated 3 days ago โ€ข 34.8k โ€ข 46 OpenMOSS-Team/MOVA-360p Image-to-Video โ€ข Updated 6 days ago โ€ข 22.7k โ€ข 200 lm-provers/QED-Nano Text Generation โ€ข 4B โ€ข Updated 5 days ago โ€ข 450 โ€ข 74 Qwen/Qwen3-ASR-1.7B Automatic Speech Recognition โ€ข Updated 22 days ago โ€ข 423k โ€ข 514 DMindAI/DMind-3 Text Generation โ€ข Updated 25 days ago โ€ข 322 โ€ข 85
========================================
[SOURCE: https://huggingface.co/datasets?modality=modality%3Ageospatial] | [TOKENS: 781]
Datasets blanchon/AID Preview โ€ข Updated Dec 4, 2023 โ€ข 719 โ€ข 8 blanchon/ETCI-2021-Flood-Detection Preview โ€ข Updated Dec 4, 2023 โ€ข 902 โ€ข 12 RichardErkhov/DASP Updated Mar 9, 2025 โ€ข 40.6k โ€ข 4 AdaptLLM/remote-sensing-VQA-benchmark Viewer โ€ข Updated Aug 21, 2025 โ€ข 24.8k โ€ข 83 โ€ข 4 nyuuzyou/streetview Preview โ€ข Updated 18 days ago โ€ข 33 โ€ข 1 hjvsl/GeoZero_Train_Datasets Viewer โ€ข Updated 9 days ago โ€ข 291k โ€ข 68 โ€ข 2 cimadure/snow_removal_transactions_in_montreal Preview โ€ข Updated May 16, 2023 โ€ข 34 โ€ข 3 kraina/airbnb Viewer โ€ข Updated Jun 3, 2023 โ€ข 103k โ€ข 62 โ€ข 11 unisaacarroyov/focus_investigaciones Preview โ€ข Updated Jun 23, 2023 โ€ข 14 psalama/NYC_sensitive_sites Viewer โ€ข Updated Jun 24, 2023 โ€ข 2.04k โ€ข 34 yachay/text_coordinates_seasons Viewer โ€ข Updated Sep 22, 2023 โ€ข 624k โ€ข 54 โ€ข 2 yachay/text_coordinates_regions Viewer โ€ข Updated Sep 21, 2023 โ€ข 615k โ€ข 46 โ€ข 8 Jerry-Master/lung-tumour-study Preview โ€ข Updated Mar 28, 2024 โ€ข 367 โ€ข 1 nasa-cisto-data-science-group/tutorial-senegal-lcluc Viewer โ€ข Updated Oct 9, 2023 โ€ข 2 โ€ข 30 satellite-image-deep-learning/SODA-A Preview โ€ข Updated Oct 22, 2023 โ€ข 36 โ€ข 17 satellite-image-deep-learning/DOTAv2 Viewer โ€ข Updated Oct 28, 2023 โ€ข 2.81k โ€ข 159 โ€ข 16 jfloresf/demo Viewer โ€ข Updated Nov 12, 2023 โ€ข 45 โ€ข 360 danaroth/cuprite Viewer โ€ข Updated Nov 10, 2023 โ€ข 1 โ€ข 143 โ€ข 1 danaroth/chikusei Viewer โ€ข Updated Nov 9, 2023 โ€ข 29 โ€ข 189 โ€ข 1 danaroth/jasper_ridge Viewer โ€ข Updated Nov 10, 2023 โ€ข 1 โ€ข 194 danaroth/urban Viewer โ€ข Updated Nov 10, 2023 โ€ข 1 โ€ข 421 โ€ข 3 danaroth/samson Viewer โ€ข Updated Nov 10, 2023 โ€ข 1 โ€ข 250 jfloresf/mlstac-demo Viewer โ€ข Updated Nov 13, 2023 โ€ข 17 โ€ข 200 joshuasundance/govgis_nov2023 Viewer โ€ข Updated Nov 17, 2023 โ€ข 510k โ€ข 40 โ€ข 4 fadingNA/CovidCases.csv Preview โ€ข Updated Nov 16, 2023 โ€ข 5 danaroth/moffett_field Viewer โ€ข Updated Nov 17, 2023 โ€ข 4 โ€ข 268 IGNF/TreeSatAI-Time-Series Preview โ€ข Updated Aug 19, 2025 โ€ข 338 โ€ข 8 joshuasundance/govgis_nov2023-slim-spatial Preview โ€ข Updated Nov 23, 2023 โ€ข 29 โ€ข 1 lauransotomayor/eco_composition Viewer โ€ข Updated Jul 4, 2024 โ€ข 10 โ€ข 97 blanchon/INRIA-Aerial-Image-Labeling Preview โ€ข Updated Dec 4, 2023 โ€ข 1.46k โ€ข 17
========================================
[SOURCE: https://en.wikipedia.org/wiki/File:P_countries-vector.svg] | [TOKENS: 128]
File:P countries-vector.svg Summary Licensing File history Click on a date/time to view the file as it appeared at that time. File usage More than 100 pages use this file. The following list shows the first 100 pages that use this file only. A full list is available. View more links to this file. Global file usage The following other wikis use this file: View more global usage of this file. Metadata This file contains additional information, probably added from the digital camera or scanner used to create or digitize it. If the file has been modified from its original state, some details may not fully reflect the modified file.
========================================
[SOURCE: https://huggingface.co/docs/hub/spaces-gpus] | [TOKENS: 937]
Hub documentation Using GPU Spaces Hub and get access to the augmented documentation experience to get started Using GPU Spaces You can upgrade your Space to use a GPU accelerator using the Settings button in the top navigation bar of the Space. You can even request a free upgrade if you are building a cool demo for a side project! Longer-term, we would also like to expose non-GPU hardware, like HPU, IPU or TPU. If you have a specific AI hardware youโ€™d like to run on, please let us know (website at huggingface.co). As soon as your Space is running on GPU you can see which hardware itโ€™s running on directly from this badge: Hardware Specs In the following tables, you can see the Specs for the different upgrade options. Configure hardware programmatically You can programmatically configure your Space hardware using huggingface_hub. This allows for a wide range of use cases where you need to dynamically assign GPUs. Check out this guide for more details. Framework specific requirements Most Spaces should run out of the box after a GPU upgrade, but sometimes youโ€™ll need to install CUDA versions of the machine learning frameworks you use. Please, follow this guide to ensure your Space takes advantage of the improved hardware. Youโ€™ll need to install a version of PyTorch compatible with the built-in CUDA drivers. Adding the following two lines to your requirements.txt file should work: You can verify whether the installation was successful by running the following code in your app.py and checking the output in your Space logs: Many frameworks automatically use the GPU if one is available. This is the case for the Pipelines in ๐Ÿค— transformers, fastai and many others. In other cases, or if you use PyTorch directly, you may need to move your models and data to the GPU to ensure computation is done on the accelerator and not on the CPU. You can use PyTorchโ€™s .to() syntax, for example: If you use JAX, you need to specify the URL that contains CUDA compatible packages. Please, add the following lines to your requirements.txt file: After that, you can verify the installation by printing the output from the following code and checking it in your Space logs. The default tensorflow installation should recognize the CUDA device. Just add tensorflow to your requirements.txt file and use the following code in your app.py to verify in your Space logs. Billing Billing on Spaces is based on hardware usage and is computed by the minute: you get charged for every minute the Space runs on the requested hardware, regardless of whether the Space is used. During a Spaceโ€™s lifecycle, it is only billed when the Space is Starting or Running. This means that there is no cost during build. If a running Space starts to fail, it will be automatically suspended and the billing will stop. Spaces running on free hardware are suspended automatically if they are not used for an extended period of time (e.g. two days). Upgraded Spaces run indefinitely by default, even if there is no usage. You can change this behavior by setting a custom โ€œsleep timeโ€ in the Spaceโ€™s settings. To interrupt the billing on your Space, you can change the Hardware to CPU basic, or pause it. Additional information about billing can be found in the dedicated Hub-wide section. Do you have an awesome Space but need help covering the GPU hardware upgrade costs? We love helping out those with an innovative Space so please feel free to apply for a community GPU grant and see if yours makes the cut! This application can be found in your Space hardware repo settings in the lower left corner under โ€œsleep time settingsโ€: Set a custom sleep time If your Space runs on the default cpu-basic hardware, it will go to sleep if inactive for more than a set time (currently, 48 hours). Anyone visiting your Space will restart it automatically. If you want your Space never to deactivate or if you want to set a custom sleep time, you need to upgrade to paid hardware. By default, an upgraded Space will never go to sleep. However, you can use this setting for your upgraded Space to become idle (stopped stage) when itโ€™s unused ๐Ÿ˜ด. You are not going to be charged for the upgraded hardware while it is asleep. The Space will โ€˜wake upโ€™ or get restarted once it receives a new visitor. The following interface will then be available in your Spaces hardware settings: The following options are available: Pausing a Space You can pause a Space from the repo settings. A โ€œpausedโ€ Space means that the Space is on hold and will not use resources until manually restarted, and only the owner of a paused Space can restart it. Paused time is not billed.
========================================
[SOURCE: https://huggingface.co/datasets?modality=modality%3Aimage] | [TOKENS: 849]
Datasets ma-xu/fine-t2i Viewer โ€ข Updated 1 day ago โ€ข 727k โ€ข 28.2k โ€ข 85 commoncrawl/CommonLID Viewer โ€ข Updated 12 days ago โ€ข 373k โ€ข 207 โ€ข 31 deepgenteam/DeepGen-1.0 Viewer โ€ข Updated 8 days ago โ€ข 1 โ€ข 1.29k โ€ข 26 markov-ai/computer-use Viewer โ€ข Updated 8 days ago โ€ข 313 โ€ข 130 โ€ข 19 nyuuzyou/suno Viewer โ€ข Updated 18 days ago โ€ข 660k โ€ข 374 โ€ข 132 cais/hle Benchmark โ€ข Updated Jan 20 โ€ข 2.5k โ€ข 32.7k โ€ข 709 kensho/PubTables-v2 Viewer โ€ข Updated 9 days ago โ€ข 3M โ€ข 156 โ€ข 12 nvidia/PhysicalAI-Robotics-NuRec Preview โ€ข Updated 2 days ago โ€ข 847 โ€ข 47 VLR-CVC/DocVQA-2026 Viewer โ€ข Updated 1 day ago โ€ข 25 โ€ข 234 โ€ข 7 ILSVRC/imagenet-1k Viewer โ€ข Updated Sep 17, 2025 โ€ข 1.43M โ€ข 95.7k โ€ข 736 none-yet/anime-captions Viewer โ€ข Updated Nov 18, 2025 โ€ข 337k โ€ข 1.05k โ€ข 28 MMMU/MMMU Viewer โ€ข Updated 9 days ago โ€ข 11.6k โ€ข 74.5k โ€ข 319 OmniAICreator/ASMR-Archive-Processed Viewer โ€ข Updated 6 days ago โ€ข 18.6M โ€ข 44k โ€ข 77 mercor/apex-agents Viewer โ€ข Updated about 12 hours ago โ€ข 480 โ€ข 16.1k โ€ข 92 microsoft/WebSTAR Preview โ€ข Updated 12 days ago โ€ข 150 โ€ข 9 moonworks/lunara-aesthetic-image-variations Viewer โ€ข Updated 12 days ago โ€ข 2.85k โ€ข 6.94k โ€ข 62 jtatman/stable-diffusion-prompts-stats-full-uncensored Viewer โ€ข Updated Nov 8, 2024 โ€ข 897k โ€ข 359 โ€ข 120 ARTPARK-IISc/Vaani Viewer โ€ข Updated Dec 22, 2025 โ€ข 22.8M โ€ข 5.77k โ€ข 102 bitmind/Nano-banana-150k Preview โ€ข Updated Oct 13, 2025 โ€ข 76 โ€ข 6 moonshotai/WorldVQA Viewer โ€ข Updated 17 days ago โ€ข 3k โ€ข 6.46k โ€ข 63 huggingface/documentation-images Viewer โ€ข Updated 2 days ago โ€ข 59 โ€ข 1.9M โ€ข 109 gaia-benchmark/GAIA Viewer โ€ข Updated Oct 28, 2025 โ€ข 932 โ€ข 14.2k โ€ข 614 Hothan/OlympiadBench Viewer โ€ข Updated Jun 8, 2025 โ€ข 8.48k โ€ข 2.6k โ€ข 41 histai/SPIDER-breast Viewer โ€ข Updated Apr 7, 2025 โ€ข 985k โ€ข 70 โ€ข 8 UCSC-VLAA/GPT-Image-Edit-1.5M Viewer โ€ข Updated Aug 21, 2025 โ€ข 2.78M โ€ข 3.49k โ€ข 72 janok24/NSFWinspo Viewer โ€ข Updated Oct 24, 2025 โ€ข 16 โ€ข 30 โ€ข 4 kaupane/nano-banana-pro-gen Viewer โ€ข Updated 5 days ago โ€ข 1k โ€ข 95 โ€ข 6 X779/ChenkinNoobXL-Style-test Viewer โ€ข Updated 8 days ago โ€ข 1 โ€ข 306 โ€ข 5 davanstrien/enc-brit-glm-ocr-v2-full Viewer โ€ข Updated 3 days ago โ€ข 2.72k โ€ข 185 โ€ข 3 zalando-datasets/fashion_mnist Viewer โ€ข Updated Aug 8, 2024 โ€ข 70k โ€ข 19.6k โ€ข 66
========================================
[SOURCE: https://huggingface.co/models?pipeline_tag=text-to-image] | [TOKENS: 971]
Models deepgenteam/DeepGen-1.0 Text-to-Image โ€ข Updated about 8 hours ago โ€ข 24 โ€ข 127 shallowdream204/BitDance-14B-16x Text-to-Image โ€ข 15B โ€ข Updated 3 days ago โ€ข 117 โ€ข 59 shallowdream204/BitDance-14B-64x Text-to-Image โ€ข 15B โ€ข Updated 3 days ago โ€ข 226 โ€ข 35 Tongyi-MAI/Z-Image-Turbo Text-to-Image โ€ข Updated 22 days ago โ€ข 990k โ€ข โ€ข 4.14k Phr00t/Qwen-Image-Edit-Rapid-AIO Text-to-Image โ€ข Updated 18 days ago โ€ข 1.65k stabilityai/stable-diffusion-xl-base-1.0 Text-to-Image โ€ข Updated Oct 30, 2023 โ€ข 2.09M โ€ข โ€ข 7.46k Tongyi-MAI/Z-Image Text-to-Image โ€ข Updated 24 days ago โ€ข 31.9k โ€ข โ€ข 930 black-forest-labs/FLUX.1-dev Text-to-Image โ€ข Updated Jun 27, 2025 โ€ข 687k โ€ข โ€ข 12.3k kpsss34/FHDR_Uncensored Text-to-Image โ€ข 12B โ€ข Updated Nov 19, 2025 โ€ข 11.8k โ€ข 135 black-forest-labs/FLUX.1-schnell Text-to-Image โ€ข Updated Aug 16, 2024 โ€ข 676k โ€ข โ€ข 4.62k Qwen/Qwen-Image-2512 Text-to-Image โ€ข Updated Dec 31, 2025 โ€ข 83.4k โ€ข โ€ข 677 lodestones/Chroma2-Kaleidoscope Text-to-Image โ€ข Updated 8 days ago โ€ข 99 alibaba-pai/Z-Image-Fun-Lora-Distill Text-to-Image โ€ข Updated 9 days ago โ€ข 9.22k โ€ข 75 Qwen/Qwen-Image Text-to-Image โ€ข Updated Aug 18, 2025 โ€ข 148k โ€ข โ€ข 2.4k MaxedOut/ComfyUI-Starter-Packs Text-to-Image โ€ข 5B โ€ข Updated 2 days ago โ€ข 3.17k โ€ข 143 Nurburgring/BEYOND_REALITY_Z_IMAGE Text-to-Image โ€ข Updated 11 days ago โ€ข 18.6k โ€ข 117 RomixERR/Pornmaster_v1-Z-Images-Turbo Text-to-Image โ€ข Updated Dec 25, 2025 โ€ข 5.93k โ€ข โ€ข 39 stepfun-ai/NextStep-1.1-Pretrain-256px Text-to-Image โ€ข 15B โ€ข Updated 5 days ago โ€ข 18 โ€ข 10 wikeeyang/Flux2-Klein-9B-True-V1 Text-to-Image โ€ข 9B โ€ข Updated 24 days ago โ€ข 8.36k โ€ข 71 unsloth/Z-Image-GGUF Text-to-Image โ€ข 6B โ€ข Updated 24 days ago โ€ข 30.3k โ€ข 112 GuangyuanSD/Z-Image-Distilled Text-to-Image โ€ข Updated 2 days ago โ€ข 6.79k โ€ข 101 ReCodePlus/Smnth_v1_NSFW1 Text-to-Image โ€ข Updated 20 days ago โ€ข 726 โ€ข โ€ข 8 CodeGoat24/FLUX.2-klein-base-9B-UnifiedReward-Flex-lora Text-to-Image โ€ข Updated 13 days ago โ€ข 377 โ€ข 17 GuangyuanSD/FLUX.2-klein-9B-Blitz-ComfyUI Text-to-Image โ€ข Updated 5 days ago โ€ข 948 โ€ข 8 xinsir/controlnet-union-sdxl-1.0 Text-to-Image โ€ข Updated Jul 30, 2024 โ€ข 179k โ€ข 1.68k stabilityai/stable-diffusion-3.5-medium Text-to-Image โ€ข Updated Oct 31, 2024 โ€ข 117k โ€ข โ€ข 905 Heartsync/Flux-NSFW-uncensored Text-to-Image โ€ข Updated May 5, 2025 โ€ข 291 nphSi/Z-Image-Lora Text-to-Image โ€ข Updated about 11 hours ago โ€ข 62.3k โ€ข 31 lodestones/Zeta-Chroma Text-to-Image โ€ข Updated 5 minutes ago โ€ข 122 BiliSakura/BitDance-14B-64x-diffusers Text-to-Image โ€ข Updated 3 days ago โ€ข 233 โ€ข 5
========================================
[SOURCE: https://www.lomdimbareshet.net/%d7%9e%d7%aa%d7%9e%d7%98%d7%99%d7%a7%d7%94-%d7%9b%d7%99%d7%aa%d7%94-%d7%98/] | [TOKENS: 1110]
ืžืชืžื˜ื™ืงื” ืœื›ื™ืชื” ื˜' ืชื•ื›ื ื™ืช ื”ืœื™ืžื•ื“ื™ื ื‘ืžืชืžื˜ื™ืงื” ืœื›ื™ืชื” ื˜' ืžื•ืจื›ื‘ืช ืžืฉืœื•ืฉื” ืชื—ื•ืžื™ื: ืืœื’ื‘ืจื™ , ื’ืื•ืžื˜ืจื™ ื•ืžืกืคืจื™ ( ื—ื™ืฉื•ื‘ื™ ).ืœืคื ื™ื›ื ื ื•ืฉืื™ ื”ืชื•ื›ื ื™ืช ื‘ืชื•ืกืคืช ืกืจื˜ื•ื ื™ ื”ืกื‘ืจ ืœื—ื•ืžืจ ื”ืœื™ืžื•ื“. ื‘ืชื—ืชื™ืช ื”ื“ืฃ ืชืžืฆืื• ืžื‘ื—ื ื™ ืžืคืž"ืจ ืžืฉื ื™ื ืงื•ื“ืžื•ืช. ืœืžื™ื“ื” ืžื”ื ื”. ื ื•ืกื—ืื•ืช ื”ื›ืคืœ ื”ืžืงื•ืฆืจ ื•ืคื™ืจื•ืง ืœื’ื•ืจืžื™ื ืคืชืจื•ืŸ ืžืฉื•ื•ืื•ืช ืจื™ื‘ื•ืขื™ื•ืช ืคืชืจื•ืŸ ืžืฉื•ื•ืื•ืช ืจื™ื‘ื•ืขื™ื•ืช ื‘ืืžืฆืขื•ืช ื”ืฉืœืžื” ืœืจื™ื‘ื•ืข ืคื™ืจื•ืง ื˜ืจื™ื ื•ื ื™ื™ืฆื•ื’ื™ื ืฉื•ื ื™ื ืœืคื•ื ืงืฆื™ื” ื”ืจื™ื‘ื•ืขื™ืช ืฉืืœื•ืช ืžื™ืœื•ืœื™ื•ืช โ€“ ืžืขืœื” ืฉื ื™ื™ื” ืื™ ืฉื™ื•ื•ื™ื•ื ื•ืช ืจื™ื‘ื•ืขื™ื™ื ืฉืืœื•ืช ืžื™ืœื•ืœื™ื•ืช ืขื ืฉื ื™ ื ืขืœืžื™ื ืžืขืจื›ืช ืžืฉื•ื•ืื•ืช ืขื ืฉื ื™ ื ืขืœืžื™ื โ€“ ืžืขืœื” ืฉื ื™ื™ื” ืฉืืœื•ืช ืžื™ืœื•ืœื™ื•ืช โ€“ ืฉื ื™ ื ืขืœืžื™ื โ€“ ืžืขืœื” ืฉื ื™ื™ื” ื”ื›ืจื•ืช ืขื ืคื•ื ืงืฆื™ื•ืช ื—ื•ืงื™ ื—ื–ืงื•ืช ืฉื•ืจืฉื™ื ืจื™ื‘ื•ืขื™ื™ื ื”ืกืชื‘ืจื•ืช ืžื•ืชื ื™ืช ื”ืกืชื‘ืจื•ืช ืฉืœ ืฉื ื™ ืžืื•ืจืขื•ืช ืžืจื•ื‘ืขื™ื ืžืฉื•ืœืฉื™ื ื‘ื ื™ื•ืช ื’ืื•ืžื˜ืจื™ื•ืช ืžืขื’ืœื™ื ืžืจื—ื‘ ืขื‘ื•ืจ ื”ืžืชื›ื•ื ื ื™ื ืœื‘ื—ื™ื ืช ื”ืกื™ื›ื•ื / ืžืคืž"ืจ , ืžืฆื•ืจืคื™ื ืžื‘ื—ื ื™ื ืžื”ืฉื ื™ื ื”ืื—ืจื•ื ื•ืช .ื”ืžื‘ื—ื ื™ื ืžื—ื•ืœืงื™ื ืœืคื™ ืจืžื•ืช ืงื•ืฉื™. ืžื‘ื—ื ื™ ื”ืกื™ื›ื•ื ื›ื•ืœืœื™ื ืฉืืœื•ืช ืžืชื•ืš ื›ืœ ืชื•ื›ื ื™ืช ื”ืœื™ืžื•ื“ื™ื ื•ืœื›ืŸ ืžื•ืžืœืฅ ืœื‘ืฆืขื ื‘ืกื•ืฃ ื›ื™ืชื” ื˜' . ืœื ื•ื—ื™ื•ืชื›ื ืžืฆื•ืจืฃ ืคืชืจื•ืŸ ืžื‘ื—ืŸ ืžืคืž"ืจ ืžืฉื ืช 2019 . ืชืจื’ื™ืœื™ ื—ื–ืจื” ืœื›ื™ืชื” ื˜' ืœืกื™ื›ื•ื ื—ื•ืžืจ ื”ืœื™ืžื•ื“ ื‘ื—ื˜ื™ื‘ืช ื”ื‘ื™ื ื™ื™ื , ืžืฆื•ืจืคื™ื ื“ืคื™ ืชืจื’ื•ืœ ืžื—ื•ืœืงื™ื ืœืคื™ ืจืžื•ืช ืœื™ืžื•ื“.ื“ืคื™ ื”ืชืจื’ื•ืœ ื”ื ื‘ื ื•ืฉืื™ื ืฉื•ื ื™ื ื•ืžื•ืžืœืฆื™ื ืœืชืจื’ื•ืœ ื•ื—ื–ืจื” ืœืคื ื™ ืฉื ื•ืช ื”ืœื™ืžื•ื“ ื‘ืชื™ื›ื•ืŸ. ืฆืจื• ืงืฉืจ - ื ืฉืžื— ืœืขื–ื•ืจ. ยฉ 2024 ื›ืœ ื”ื–ื›ื•ื™ื•ืช ืฉืžื•ืจื•ืช. ืœื•ืžื“ื™ื ื‘ืจืฉืช - ื“ื•ื“ื• ื’ื•ืœื“ืฉื˜ื™ื™ืŸ ืขืจื•ืฅ "ืœื•ืžื“ื™ื ื‘ืจืฉืช" ืœื™ืžื•ื“ ืขืฆืžื™
========================================
[SOURCE: https://huggingface.co/models?language=en] | [TOKENS: 906]
Models nvidia/personaplex-7b-v1 Audio-to-Audio โ€ข Updated 6 days ago โ€ข 539k โ€ข 2.1k zai-org/GLM-5 Text Generation โ€ข 754B โ€ข Updated 8 days ago โ€ข 177k โ€ข โ€ข 1.39k Nanbeige/Nanbeige4.1-3B Text Generation โ€ข 4B โ€ข Updated 2 days ago โ€ข 130k โ€ข โ€ข 636 FireRedTeam/FireRed-Image-Edit-1.0 Image-to-Image โ€ข Updated 7 days ago โ€ข 2.15k โ€ข โ€ข 228 nineninesix/kani-tts-2-en Text-to-Speech โ€ข 0.4B โ€ข Updated 2 days ago โ€ข 2.59k โ€ข 163 OpenMOSS-Team/MOSS-TTS Text-to-Speech โ€ข 8B โ€ข Updated 8 days ago โ€ข 41.2k โ€ข 289 jdopensource/JoyAI-LLM-Flash Text Generation โ€ข 49B โ€ข Updated 3 days ago โ€ข 807 โ€ข 128 nvidia/NVIDIA-Nemotron-Nano-9B-v2-Japanese Text Generation โ€ข 9B โ€ข Updated 1 day ago โ€ข 5.14k โ€ข 92 mistralai/Voxtral-Mini-4B-Realtime-2602 Automatic Speech Recognition โ€ข Updated 3 days ago โ€ข 101k โ€ข 596 Zyphra/ZUNA 0.4B โ€ข Updated 3 days ago โ€ข 747 โ€ข 84 Soul-AILab/SoulX-Singer Text-to-Speech โ€ข Updated 10 days ago โ€ข 1.08k โ€ข 125 zai-org/GLM-OCR Image-to-Text โ€ข Updated 12 days ago โ€ข 1.32M โ€ข 1.1k shallowdream204/BitDance-14B-16x Text-to-Image โ€ข 15B โ€ข Updated 3 days ago โ€ข 117 โ€ข 59 unsloth/GLM-5-GGUF Text Generation โ€ข 754B โ€ข Updated 6 days ago โ€ข 48.1k โ€ข 182 salakash/Minimalism Text Generation โ€ข Updated about 20 hours ago โ€ข 1.19k โ€ข 76 inclusionAI/Ming-flash-omni-2.0 Any-to-Any โ€ข Updated 9 days ago โ€ข 7.86k โ€ข 241 CohereLabs/tiny-aya-global Text Generation โ€ข 3B โ€ข Updated 2 days ago โ€ข 988 โ€ข 47 openbmb/MiniCPM-SALA Text Generation โ€ข Updated 10 days ago โ€ข 4.66k โ€ข 470 lm-provers/QED-Nano Text Generation โ€ข 4B โ€ข Updated 5 days ago โ€ข 450 โ€ข 74 unsloth/GLM-4.7-Flash-GGUF Text Generation โ€ข 30B โ€ข Updated 9 days ago โ€ข 360k โ€ข 502 DMindAI/DMind-3 Text Generation โ€ข Updated 25 days ago โ€ข 322 โ€ข 85 zai-org/GLM-4.7-Flash Text Generation โ€ข Updated 23 days ago โ€ข 1.75M โ€ข โ€ข 1.55k hexgrad/Kokoro-82M Text-to-Speech โ€ข Updated Apr 10, 2025 โ€ข 8.66M โ€ข โ€ข 5.73k zai-org/GLM-5-FP8 Text Generation โ€ข Updated 8 days ago โ€ข 421k โ€ข 113 Lightricks/LTX-2 Image-to-Video โ€ข Updated 18 days ago โ€ข 1.97M โ€ข โ€ข 1.56k mmnga-o/NVIDIA-Nemotron-Nano-9B-v2-Japanese-gguf Text Generation โ€ข 9B โ€ข Updated 3 days ago โ€ข 6.16k โ€ข 36 Aratako/MioTTS-2.6B Text-to-Speech โ€ข Updated 11 days ago โ€ข 924 โ€ข 63 shallowdream204/BitDance-14B-64x Text-to-Image โ€ข 15B โ€ข Updated 3 days ago โ€ข 226 โ€ข 35 Tongyi-MAI/Z-Image-Turbo Text-to-Image โ€ข Updated 22 days ago โ€ข 990k โ€ข โ€ข 4.14k CohereLabs/tiny-aya-base Text Generation โ€ข 3B โ€ข Updated 2 days ago โ€ข 467 โ€ข 34
========================================
[SOURCE: https://he.wikipedia.org/wiki/ืจื“ื™ืคื”] | [TOKENS: 16680]
ืชื•ื›ืŸ ืขื ื™ื™ื ื™ื ืจื“ื™ืคื” ืจื“ื™ืคื” ื”ื™ื ื”ืชื™ื™ื—ืกื•ืช ืžืคืœื” ื‘ืื•ืคืŸ ืฉื™ื˜ืชื™ ื›ืœืคื™ ืื“ื ืื• ืงื‘ื•ืฆื” ืขืœ ื™ื“ื™ ืื“ื ืื• ืงื‘ื•ืฆื” ืื—ืจืช. ื”ืฆื•ืจื•ืช ื”ื ืคื•ืฆื•ืช ื‘ื™ื•ืชืจ ื”ืŸ ืจื“ื™ืคื•ืช ื“ืชื™ื•ืช, ื’ื–ืขื ื•ืช (ืจื“ื™ืคื” ืืชื ื™ืช), ืจื“ื™ืคื•ืช ืคื•ืœื™ื˜ื™ื•ืช ืื ื›ื™ ื‘ืื•ืคืŸ ื˜ื‘ืขื™ ื™ืฉ ื—ืคื™ืคื” ืžืกื•ื™ืžืช ื‘ื™ืŸ ืžื•ื ื—ื™ื ืืœื”. ื›ืžื• ื›ืŸ ืงื™ื™ืžืช ืจื“ื™ืคื” ืขืœ ื‘ืกื™ืก ื ื˜ื™ื™ื” ืžื™ื ื™ืช ืื• ืžื’ื“ืจื™ืช โ€“ ื”ื•ืžื•ืคื•ื‘ื™ื” ื•ื˜ืจื ืกืคื•ื‘ื™ื” ื•ื›ืŸ ืจื“ื™ืคื” ื‘ืฉืœ ืฉื•ื ื™ ื‘ืžืจืื” ืื• ื‘ืขื™ื•ืช ื‘ืจื™ืื•ืช ื›ืžื• ืจื“ื™ืคื” ืœื‘ืงื ื™ื ืื• ืฉืœ ืื ืฉื™ื ื‘ืขืœื™ ืžื—ืœื•ืช ื ืคืฉ. ื”ื’ื“ืจืช ืจื“ื™ืคื” ื›ื•ืœืœืช ื”ื˜ืœืช ืกื‘ืœ, ื”ื˜ืจื“ื”, ืžืืกืจ, ืคื—ื“ ืื• ื›ืื‘ ื•ืœืคืขืžื™ื ื”ืจื“ื™ืคื” ืžื’ื™ืขื” ืœื™ื“ื™ ืžืขืฉื™ ื˜ื‘ื—, ื˜ื™ื”ื•ืจ ืืชื ื™, ืื• ืจืฆื— ืขื. ืขื ื–ืืช ืœื ื›ืœ ื’ืจื™ืžืช ืกื‘ืœ ืžื•ื’ื“ืจืช ื‘ื”ื›ืจื— ื›ืจื“ื™ืคื”. ื”ืกื‘ืœ ืฉื—ื•ื•ื” ื”ืงื•ืจื‘ืŸ ื—ื™ื™ื‘ ืœื”ื™ื•ืช ื—ืžื•ืจ ื‘ืžื™ื“ื” ืžืกื•ื™ืžืช. ืงื™ื™ืžืช ืžื—ืœื•ืงืช ืœื’ื‘ื™ ืžื™ื“ืช ื—ื•ืžืจืชื• ืฉืœ ื”ืกื‘ืœ, ื—ื•ืžืจื” ืฉืžื’ื“ื™ืจื” ืจื“ื™ืคื”. ื—ื•ืง ื‘ื™ื ืœืื•ืžื™ ื›ื—ืœืง ืžืขืงืจื•ื ื•ืช ื ื™ืจื ื‘ืจื’, ืคืฉืขื™ื ื ื’ื“ ื”ืื ื•ืฉื•ืช ื”ื ื—ืœืง ืžื”ืžืฉืคื˜ ื”ื‘ื™ื ืœืื•ืžื™. ืขืงืจื•ืŸ VI ืฉืœ ืขืงืจื•ื ื•ืช ื ื™ืจื ื‘ืจื’ ืงื•ื‘ืข ื›ื™ ื”ืคืฉืขื™ื ื”ืžืคื•ืจื˜ื™ื ืœื”ืœืŸ ื“ื™ื ื ื›ืขื‘ื™ืจื•ืช ืœืคื™ ื”ื—ื•ืง ื”ื‘ื™ื ืœืื•ืžื™: ... (ื’) ืคืฉืขื™ื ื ื’ื“ ื”ืื ื•ืฉื•ืช: ื˜ืœืคื•ืจื“ ื˜ื™ื™ืœื•ืจ, ืฉื”ื™ื” ื™ื•ืขืฅ ื”ืชื‘ื™ืขื” ื‘ืžืฉืคื˜ื™ ื ื™ืจื ื‘ืจื’ ื›ืชื‘ "ื‘ืžืฉืคื˜ ืคืฉืขื™ ื”ืžืœื—ืžื” ื ื™ืจื ื‘ืจื’, ืฉืœืœื• ื‘ืชื™ ื”ืžืฉืคื˜ ืืช ื”ื”ื’ื“ืจื” ืฉื‘ื™ืงืฉื” ื”ืชื‘ื™ืขื” ืœืืžืฅ ืขื‘ื•ืจ ืžืขืฉื™ ื–ื•ื•ืขื” "ืคื ื™ื-ืžื“ื™ื ืชื™ื™ื" ื›"ืคืฉืขื™ื ื ื’ื“ ื”ืื ื•ืฉื•ืช" ื‘ืžืกื’ืจืช ื”ืžืฉืคื˜ ื”ื‘ื™ื ืœืื•ืžื™. ื‘ื›ืžื” ืืžื ื•ืช ื‘ื™ื ืœืื•ืžื™ื•ืช ืžืื•ื—ืจื•ืช ื™ื•ืชืจ ืžืฉื•ืœื‘ ืขื™ืงืจื•ืŸ ื–ื”, ืืš ื‘ื—ืœืงืŸ ืžื•ืฉืžื˜ืช ื”ื”ื’ื“ืจื” "ื‘ืงืฉืจ ืœืคืฉืข ื›ืœืฉื”ื• ื ื’ื“ ื”ืฉืœื•ื ืื• ืคืฉืข ืžืœื—ืžื” ื›ืœืฉื”ื•", ื”ื’ื“ืจื” ืฉื ืžืฆืืช ื‘ืขืงืจื•ื ื•ืช ื ื™ืจื ื‘ืจื’. ืืžื ืช ืจื•ืžื ืฉืœ ื‘ื™ืช ื”ื“ื™ืŸ ื”ืคืœื™ืœื™ ื”ื‘ื™ื ืœืื•ืžื™, ื”ืžื—ื™ื™ื‘ืช 111 ืžื“ื™ื ื•ืช, ืžื’ื“ื™ืจื” ืคืฉืขื™ื ื ื’ื“ ื”ืื ื•ืฉื•ืช ื‘ืกืขื™ืฃ 7.1. ื”ืกืขื™ืฃ ืžื’ื“ื™ืจ ื›ืžืขืฉื™ ืคืฉืข ืžืขืฉื™ื ืžืกื•ื™ืžื™ื "ืฉื ืขืฉื• ื‘ืžืกื’ืจืช ืžืชืงืคื” ืจื—ื‘ื” ืื• ืฉื™ื˜ืชื™ืช ื”ืžื•ืคื ื™ืช ื ื’ื“ ื›ืœ ืื•ื›ืœื•ืกื™ื™ื” ืื–ืจื—ื™ืช, ืžืชื•ืš ืžื•ื“ืขื•ืช ืœืคื’ื™ืขื”". ืืœื• ื›ื•ืœืœื™ื: (ื—) ืจื“ื™ืคื” ื›ื ื’ื“ ื›ืœ ืงื‘ื•ืฆื” ืื• ืงื•ืœืงื˜ื™ื‘ ืฉื ื™ืชืŸ ืœื–ื”ื•ืชื ืขืœ ืจืงืข ืคื•ืœื™ื˜ื™, ื’ื–ืขื™, ืœืื•ืžื™, ืืชื ื™, ืชืจื‘ื•ืชื™, ื“ืชื™, ืžื’ื“ืจื™. ืื• ืžื˜ืขืžื™ื ืื—ืจื™ื ื”ืžื•ื›ืจื™ื ื‘ืื•ืคืŸ ืื•ื ื™ื‘ืจืกืœื™ ื›ื‘ืœืชื™ ืžื•ืชืจื™ื ืขืœ ืคื™ ื”ื—ื•ืง ื”ื‘ื™ื ืœืื•ืžื™, ื‘ืงืฉืจ ืœื›ืœ ืžืขืฉื” ื”ืžื•ื–ื›ืจ ื‘ืคืกืงื” ื–ื• (ืœืžืฉืœ ืจืฆื—, ื”ืฉืžื“ื”, ืฉืขื‘ื•ื“, ื’ื™ืจื•ืฉ, ืžืืกืจ, ืขื™ื ื•ื™ื™ื, ืืœื™ืžื•ืช ืžื™ื ื™ืช, ืืคืจื˜ื”ื™ื™ื“ ื•ืžืขืฉื™ื ืœื ืื ื•ืฉื™ื™ื ืื—ืจื™ื) ืื• ื›ืœ ืคืฉืข ื‘ืกืžื›ื•ืช ื‘ื™ืช ื”ืžืฉืคื˜. ืจื“ื™ืคื” ืขืœ ืจืงืข ื“ืชื™ ืจื“ื™ืคื” ื“ืชื™ืช ื”ื™ื ื”ืชื™ื™ื—ืกื•ืช ืฉืœื™ืœื™ืช ืฉื™ื˜ืชื™ืช ืœืื“ื ืื• ืงื‘ื•ืฆื” ืขืงื‘ ื”ืฉืชื™ื™ื›ื•ืชื ื”ื“ืชื™ืช. ืชืื•ืจื˜ื™ืงื ื™ ื”ื—ื™ืœื•ืŸ (ื”ืžื ื™ื—ื™ื ื™ืจื™ื“ืช ื“ืชื™ื•ืช ื‘ืื•ืคืŸ ื›ืœืœื™) ืกื‘ืจื• ื‘ืชืงื•ืคื•ืช ืฉื•ื ื•ืช ืฉืจืืฉื™ืชืŸ ื‘ืขื™ื“ืŸ ื”ื ืื•ืจื•ืช, ืฉื”ืจื“ื™ืคื” ื”ื“ืชื™ืช ื”ื™ื ื ื—ืœืช ื”ืขื‘ืจ.[ื“ืจื•ืฉ ืžืงื•ืจ] ืขื ื–ืืช, ืขื ืขืœื™ื™ืช ื”ืคื•ื ื“ืžื ื˜ืœื™ื–ื ื•ื”ื˜ืจื•ืจ ื”ืงืฉื•ืจ ื‘ืื“ื™ืงื•ืช ื”ื“ืชื™ืช, ื”ื ื—ื” ื–ื• ื”ืคื›ื” ืœืฉื ื•ื™ื” ื™ื•ืชืจ ื‘ืžื—ืœื•ืงืช[ื“ืจื•ืฉ ืžืงื•ืจ] ื‘ืžื“ื™ื ื•ืช ืจื‘ื•ืช ื‘ืขื•ืœื ื›ื™ื•ื ืจื“ื™ืคื” ื“ืชื™ืช ื”ื™ื ื‘ืขื™ื” ืฉืœ ื–ื›ื•ื™ื•ืช ืื“ื. ืืชืื™ืกื˜ื™ื ื—ื•ื• ืจื“ื™ืคื•ืช ืœืื•ืจืš ื”ื”ื™ืกื˜ื•ืจื™ื”: ืžืขืฆืจ, ื›ืœื™ืื”, ื”ื›ืื”, ืขื™ื ื•ื™ื™ื ืื• ื”ื•ืฆืื” ืœื”ื•ืจื’ ืฉืœื ื›ื“ื™ืŸ, ื•ื›ืŸ ื”ื—ืจืžืช ืจื›ื•ืฉ ืื• ื”ืฉืžื“ืชื•. ืจื“ื™ืคืช ื”ื‘ื”ืื™ื ืžืชื™ื™ื—ืกืช ืœืจื“ื™ืคื” ื“ืชื™ืช ืฉืœ ื‘ื”ืื™ื ื‘ืžื“ื™ื ื•ืช ืฉื•ื ื•ืช, ื‘ืžื™ื•ื—ื“ ื‘ืื™ืจืืŸ, ื‘ื” ืื—ืช ื”ืื•ื›ืœื•ืกื™ื•ืช ื”ื‘ื”ืื™ื•ืช ื”ื’ื“ื•ืœื•ืช ื‘ืขื•ืœื. ื”ืืžื•ื ื” ื”ื‘ื”ืื™ืช ืžืงื•ืจื” ื‘ืื™ืจืืŸ ื•ื”ืžื™ืขื•ื˜ ื”ื“ืชื™ ื”ื’ื“ื•ืœ ื‘ื™ื•ืชืจ ื‘ืžื“ื™ื ื” ื–ื• ื”ื•ื ื‘ื”ืื™. ืจื“ื™ืคืช ื”ื ื•ืฆืจื™ื ื”ื™ื ืจื“ื™ืคื” ื“ืชื™ืช ืฉื ื•ืฆืจื™ื ืขืฉื•ื™ื™ื ืœืขื‘ื•ืจ ื›ืชื•ืฆืื” ืžื”ืฆื”ืจืช ืืžื•ื ืชื, ื”ืŸ ืžื‘ื—ื™ื ื” ื”ื™ืกื˜ื•ืจื™ืช ื•ื”ืŸ ื‘ืขื™ื“ืŸ ื”ื ื•ื›ื—ื™. ื ื•ืฆืจื™ื ืžื•ืงื“ืžื™ื ื ืจื“ืคื• ื‘ื’ืœืœ ืืžื•ื ืชื ื‘ื™ื“ื™ ื™ื”ื•ื“ื™ื-ื ื•ืฆืจื™ื ื•ื’ื ื‘ื™ื“ื™ ื”ืื™ืžืคืจื™ื” ื”ืจื•ืžื™ืช ืฉืฉืœื˜ื” ื‘ื—ืœืง ื ื™ื›ืจ ืžืื–ื•ืจื™ ืชืคื•ืฆืชื” ืฉืœ ื”ื ืฆืจื•ืช ื”ืงื“ื•ืžื”. ื‘ืชื—ื™ืœืช ื”ืžืื” ื”ืจื‘ื™ืขื™ืช ืื•ืฉืจื” ื–ื›ื•ืช ืงื™ื•ืžื” ืฉืœ ื”ื ืฆืจื•ืช ื‘ืื™ืžืคืจื™ื” ื”ืจื•ืžื™ืช ืขืœ ื™ื“ื™ ืฆื• ืžื™ืœืื ื•, ื•ื”ื™ื ื”ืคื›ื” ืœื“ืช ื”ืžื“ื™ื ื” ืฉืœ ื”ืื™ืžืคืจื™ื” ื”ืจื•ืžื™ืช. ืžื™ืกื™ื•ื ืจื™ื ื ื•ืฆืจื™ื, ื›ืžื• ื’ื ื”ืื ืฉื™ื ืฉื”ืžื™ืกื™ื•ื ืจื™ื ื”ืžื™ืจื• ืืช ื“ืชื ืœื ืฆืจื•ืช, ื”ื™ื• ืžื˜ืจื” ืœืจื“ื™ืคื”, ืคืขืžื™ื ืจื‘ื•ืช ืขื“ ื›ื“ื™ ืขื™ื ื•ื™ ื•ืจืฆื— ืขืœ ืžื–ื‘ื— ืืžื•ื ืชื. ื”ื™ื• ืขื“ื•ืช ื ื•ืฆืจื™ื•ืช ื‘ื•ื“ื“ื•ืช ืฉืกื‘ืœื• ืžืจื“ื™ืคื” ื‘ื™ื“ื™ ื ื•ืฆืจื™ื ืื—ืจื™ื ื”ืžื•ืคืงื“ื™ื ืขืœ ื˜ื™ืคื•ืœ ื‘ื›ื•ืคืจื™ื, ื‘ืžื™ื•ื—ื“ ื‘ืžื”ืœืš ื”ืจืคื•ืจืžืฆื™ื” ื”ืคืจื•ื˜ืกื˜ื ื˜ื™ืช ืฉืœ ื”ืžืื” ื”-16 ื•ื›ืŸ ืœืื•ืจืš ื™ืžื™ ื”ื‘ื™ื ื™ื™ื ื›ืืฉืจ ืงื‘ื•ืฆื•ืช ื ื•ืฆืจื™ื•ืช ืฉื•ื ื•ืช ืฉื ื—ืฉื‘ื• ื›ื•ืคืจื•ืช ื ืจื“ืคื• ืขืœ ื™ื“ื™ ื”ืืคื™ืคื™ื•ืจ. ื‘ืžืื” ื”-20 ื ืจื“ืคื• ื ื•ืฆืจื™ื ืขืœ ื™ื“ื™ ืงื‘ื•ืฆื•ืช ืฉื•ื ื•ืช ื•ืขืœ ื™ื“ื™ ืžื“ื™ื ื•ืช ืืชืื™ืกื˜ื™ื•ืช ื•ื˜ื•ื˜ืœื™ื˜ืจื™ื•ืช ื›ืžื• ื‘ืจื™ืช ื”ืžื•ืขืฆื•ืช ื•ืฆืคื•ืŸ ืงื•ืจื™ืื”. ื‘ืžื”ืœืš ืžืœื—ืžืช ื”ืขื•ืœื ื”ืฉื ื™ื™ื” ื ืจื“ืคื• ื—ื‘ืจื™ ื›ื ืกื™ื•ืช ื ื•ืฆืจื™ื•ืช ืจื‘ื•ืช ื‘ื’ืจืžื ื™ื” ืขืงื‘ ื”ืชื ื’ื“ื•ืชื ืœืื™ื“ืื•ืœื•ื’ื™ื” ื”ื ืืฆื™ืช, ื‘ื”ื ืงืชื•ืœื™ื ื’ืจืžื ื™ื. ื‘ืกื•ืฃ ื”ืžืื” ื”-20 ืœื”ืขืจื›ืช ืืจื’ื•ืŸ ื”ืžื™ืกื™ื•ื ืจื™ื ื”ื ื•ืฆืจื™ ื”ื‘ืจื™ื˜ื™ "ื“ืœืชื•ืช ืคืชื•ื—ื•ืช" (Oped Doors), ื”ืชืžื•ื“ื“ื• 100 ืžื™ืœื™ื•ืŸ ื ื•ืฆืจื™ื ืขื ืจื“ื™ืคื•ืช, ื‘ืžื™ื•ื—ื“ ื‘ืžื“ื™ื ื•ืช ื”ื ืฉืœื˜ื•ืช ืขืœ ื™ื“ื™ ืžื•ืกืœืžื™ื ื›ืžื• ืคืงื™ืกื˜ืŸ ื•ืขืจื‘ ื”ืกืขื•ื“ื™ืช. ืขืœ ืคื™ ื”ืื’ื•ื“ื” ื”ื‘ื™ื ืœืื•ืžื™ืช ืœื–ื›ื•ื™ื•ืช ืื“ื, ืขื“ 80% ืžื›ืœืœ ืคืขื•ืœื•ืช ื”ืจื“ื™ืคื” ื”ื“ืชื™ืช ืžื•ืคื ื•ืช ื›ื ื’ื“ ืžืืžื™ื ื™ ื”ื“ืช ื”ื ื•ืฆืจื™ืช. ืขื ืฆื• ื”ื”ืฉืžื“ื” ื‘ืžื™ื–ื•ืจื™ (ืื ') ื‘-1838 ื”ืคื›ื• ื”ืžื•ืจืžื•ื ื™ื ืœืงื‘ื•ืฆื” ื”ื“ืชื™ืช ื”ื™ื—ื™ื“ื” ืฉืืจืฆื•ืช ื”ื‘ืจื™ืช ื”ื›ืฉื™ืจื” ืืช ื”ืฉืžื“ืช ื“ืชื. ื–ื” ื”ื™ื” ืœืื—ืจ ื ืื•ื ืฉื ืฉื ืกื™ื“ื ื™ ืจื™ื’ื“ื•ืŸ ื‘-4 ื‘ื™ื•ืœื™ (ื™ื•ื ื”ืขืฆืžืื•ืช ื‘ืืจืฆื•ืช ื”ื‘ืจื™ืช), ืฉื‘ื• ืงื‘ืข ื›ื™ ื”ืžื•ืจืžื•ื ื™ื ื™ื’ื ื• ืขืœ ื—ื™ื™ื”ื ื•ืจื›ื•ืฉื. ื ืื•ื ื–ื” ื ื—ืฉื‘ ืœืžืชืกื™ืก ื•ืžืžืจื™ื“. ื’ื™ืจื•ืฉื ืฉืœ ื”ืžื•ืจืžื•ื ื™ื ื‘ื›ื•ื— ืžื”ืžื“ื™ื ื” ื’ืจื ืœืžื•ืชื ืฉืœ ืœืžืขืœื” ืžืžืื” ืขืงื‘ ื—ืฉื™ืคื”, ืจืขื‘ ื•ืžื—ืœื•ืช ื›ืชื•ืฆืื” ืžื›ืš. ื”ืžื•ืจืžื•ื ื™ื ืกื‘ืœื• ืžื–ื™ืคื•ืช ื•ื ื•ืฆื•ืช, ืื“ืžื•ืชื™ื”ื ื•ืจื›ื•ืฉื ื ืœืงื—ื• ืžื”ื ืฉื•ื‘ ื•ืฉื•ื‘, ื”ืชืงืคื•ืช ืืกืคืกื•ืฃ, ืžืืกืจื™ ืฉื•ื•ื, ื•ืืจืฆื•ืช ื”ื‘ืจื™ืช ืฉืœื—ื” ืฆื‘ื ืœื™ื•ื˜ื” ื›ื“ื™ ืœื˜ืคืœ ื‘"ื‘ืขื™ื™ืช ื”ืžื•ืจืžื•ื ื™ื " ื‘ืžืœื—ืžืช ื™ื•ื˜ื” ืฉื”ื‘ื™ืื” ืœืงื‘ื•ืฆื” ืฉืœ ืžื•ืจืžื•ื ื™ื ื‘ื”ื ื”ื’ืช ื’'ื•ืŸ ื“ 'ืœื™ ื˜ื‘ื—ื™ื ื‘ืžืชื ื—ืœื™ื ื‘ื˜ื‘ื— ื‘ื”ืจ ืžื“ื•ื–. ืžื™ืœื™ืฆื™ื” ืžืžืฉืœืชื™ืช ืฉื—ื˜ื” ื‘ืžื•ืจืžื•ื ื™ื ื‘ืžื” ืฉืžื›ื•ื ื” ื›ื™ื•ื ื˜ื‘ื— ื˜ื—ื ืช ื”ืื•ืŸ. ืžื™ื™ืกื“ ื”ื›ื ืกื™ื™ื”, ื’'ื•ื–ืฃ ืกืžื™ืช', ื ื”ืจื’ ื‘ืงืจืชื’ื•, ืื™ืœื™ื ื•ื™ ืขืœ ื™ื“ื™ ืืกืคืกื•ืฃ ืฉืœ ื›ึพ200 ืื™ืฉ, ืฉื›ืžืขื˜ ื›ื•ืœื ื”ื™ื• ื—ื‘ืจื™ื ื‘ืžื™ืœื™ืฆื™ื™ืช ืžื“ื™ื ืช ืื™ืœื™ื ื•ื™ ื›ื•ืœืœ ื›ืžื” ืžื—ื‘ืจื™ ื”ืžื™ืœื™ืฆื™ื” ืฉื”ื•ื˜ืœื• ืœืฉืžื•ืจ ืขืœื™ื•. ืขื“ื™ ื™ื”ื•ื”, ืืžื•ื ื•ืชื™ื”ื, ืชื•ืจืชื ื•ื”ืคืจืงื˜ื™ืงื•ืช ืฉืœื”ื ื”ื•ืœื™ื“ื• ืžื—ืœื•ืงืช ื•ื”ืชื ื’ื“ื•ืช ืžืฆื“ ืžืžืฉืœื•ืช ืžืงื•ืžื™ื•ืช, ืงื”ื™ืœื•ืช ื•ืงื‘ื•ืฆื•ืช ื ื•ืฆืจื™ื•ืช ืžื”ื–ืจืžื™ื ื”ืžืจื›ื–ื™ื™ื. ืขื“ื™ ื™ื”ื•ื” ื ืจื“ืคื• ื‘ืขื™ืงืจ ื‘ื’ืจืžื ื™ื” ื”ื ืืฆื™ืช. ืคืืœื•ืŸ ื’ื•ื ื’ ื”ื•ืฆื’ ื‘ืคื•ืžื‘ื™ ืœื™ ื”ื•ื ื’ื–'ื™ ื‘ืฆ'ืื ื’ืฆ'ื•ืŸ, ืกื™ืŸ, ื‘ืฉื ืช 1992. ื‘ืฉื ื™ื ืฉืœืื—ืจ ืžื›ืŸ ื”ื™ื” ื”ืคืืœื•ืŸ ื’ื•ื ื’ ื”ืชืจื’ื•ืœ ื”ืฆื•ืžื— ื‘ื™ื•ืชืจ ื‘ืฆ'ื™ ืงื•ื ื’" ื‘ื”ื™ืกื˜ื•ืจื™ื” ื”ืกื™ื ื™ืช, ื•ื‘ืฉื ืช 1999 ื”ื™ื• ืžื™ืœื™ื•ื ื™ ืžืชืจื’ืœื™ื ื‘ืฉื™ื˜ื” ื–ื•. ื‘ืขืงื‘ื•ืช ืฉื ื•ืช ื”ืคื•ืคื•ืœืจื™ื•ืช ื”ื ืจื—ื‘ืช, ื‘-20 ื‘ื™ื•ืœื™ 1999, ื”ื—ืœื” ืžืžืฉืœืช ืกื™ืŸ ื‘ืžืกืข ืจื“ื™ืคื” ืืจืฆื™ ื ื’ื“ ืžืชืจื’ืœื™ ืคืืœื•ืŸ ื’ื•ื ื’, ืœืžืขื˜ ื‘ืื–ื•ืจื™ื ื”ืžื™ื ื”ืœื™ื™ื ื”ืžื™ื•ื—ื“ื™ื ืฉืœ ื”ื•ื ื’ ืงื•ื ื’ ื•ืžืงืื•. ื‘ืกื•ืฃ 1999 ื ื—ืงืงื” ื—ืงื™ืงื” ืฉื ื•ืขื“ื” ืœื”ื•ืฆื™ื ืืœ ืžื—ื•ืฅ ืœื—ื•ืง "ื“ืชื•ืช ื”ื˜ืจื•ื“ื•ืงืกื™ื•ืช" ื•ื—ืงื™ืงื” ื–ื• ื”ื•ื—ืœื” ืจื˜ืจื•ืืงื˜ื™ื‘ื™ืช ืขืœ ื”ืคืืœื•ืŸ ื’ื•ื ื’. ืืจื’ื•ืŸ ืืžื ืกื˜ื™ ืื™ื ื˜ืจื ืฉื™ื•ื ืœ ื”ืฆื”ื™ืจ ื›ื™ ื”ืจื“ื™ืคื” "ืžื•ื ืขืช ืžืžื ื™ืขื™ื ืคื•ืœื™ื˜ื™ื™ื" ื‘ืืžืฆืขื•ืช "ื—ืงื™ืงื” ืฉืžืฉืžืฉืช ืจื˜ืจื•ืืงื˜ื™ื‘ื™ืช ืœืื ืฉื™ื ื‘ื”ืืฉืžื•ืช ื”ืžื•ื ืขื•ืช ืžืžื ื™ืขื™ื ืคื•ืœื™ื˜ื™ื™ื, ื•ื‘ืืžืฆืขื•ืช ืชืงื ื•ืช ื—ื“ืฉื•ืช ืฉื”ื•ืชืงื ื• ืœืฉื ื”ื’ื‘ืœืช ื”ื—ื™ืจื•ื™ื•ืช ื”ื‘ืกื™ืกื™ื•ืช". ืจื“ื™ืคืช ื”ื™ื ื“ื™ื ืžืชื™ื™ื—ืกืช ืœืจื“ื™ืคื” ื”ื“ืชื™ืช ืฉืขืœื•ืœื™ื ื”ื™ื ื“ื™ื ืœืขื‘ื•ืจ ืขืงื‘ ืืžื•ื ืชื, ื”ืŸ ืžื‘ื—ื™ื ื” ื”ื™ืกื˜ื•ืจื™ืช ื•ื”ืŸ ื‘ืขื™ื“ืŸ ื”ื ื•ื›ื—ื™. ื”ื™ื ื“ื™ื ื ืจื“ืคื• ื‘ืื›ื–ืจื™ื•ืช ื‘ืชืงื•ืคืช ื”ืฉืœื˜ื•ืŸ ื”ืืกืœืืžื™ ืฉืœ ืชืช ื”ื™ื‘ืฉืช ื”ื”ื•ื“ื™ืช ื•ื‘ืžื”ืœืš ื”ืฉืœื˜ื•ืŸ ื”ืคื•ืจื˜ื•ื’ืœื™ ื‘ื’ื•ืื”. ื’ื ื‘ื™ืžื™ื ื•, ื”ื™ื ื“ื™ื ื‘ืคืงื™ืกื˜ืŸ ื•ื‘ื‘ื ื’ืœื“ืฉ ืกื•ื‘ืœื™ื ืžืจื“ื™ืคื•ืช. ืืœืคื™ ื”ื™ื ื“ื™ื ืžืžื—ื•ื– ืกื™ื ื“ ื‘ืคืงื™ืกื˜ืŸ ื ืžืœื˜ื• ืœื”ื•ื“ื• ื•ื”ื‘ื™ืขื• ื—ืฉืฉ ืœืฉืœื•ืžื. ืœืื—ืจ ื—ืœื•ืงืช ื”ื•ื“ื• ื‘ืฉื ืช 1947, ื”ื™ื• ื‘ืคืงื™ืกื˜ืŸ 8.8 ืžื™ืœื™ื•ืŸ ื”ื™ื ื“ื™ื (ืœืžืขื˜ ื‘ื ื’ืœื“ืฉ) ื‘ืฉื ืช 1951 โ€“ 22% ืžืื•ื›ืœื•ืกื™ื™ืช ืคืงื™ืกื˜ืŸ (ื›ื•ืœืœ ื‘ื ื’ืœื“ืฉ ืฉืœ ื™ืžื™ื ื• ืฉื”ื™ื™ืชื” ื—ืœืง ืžืคืงื™ืกื˜ืŸ). ื›ื™ื•ื ื”ืžื™ืขื•ื˜ ื”ื”ื™ื ื“ื™ ืžืกืชื›ื ื‘ื›-1.7% ืžืื•ื›ืœื•ืกื™ื™ืช ืคืงื™ืกื˜ืŸ. ืžืœื—ืžืช ื”ืฉื—ืจื•ืจ ื‘ื‘ื ื’ืœื“ืฉ (1971) ื”ื‘ื™ืื” ืœืื—ื“ ืžืจืฆื™ื—ื•ืช ื”ืขื ื”ื’ื“ื•ืœื•ืช ื‘ืžืื” ื”-20. ืžืกืคืจ ื”ื ืคื’ืขื™ื ื ืืžื“ ื‘-3,000,000, ืืš ืกื‘ื™ืจ ืœื”ื ื™ื— ืฉื”ื”ื™ื ื“ื™ื ื ืคื’ืขื• ื‘ืื•ืคืŸ ื‘ืœืชื™ ืคืจื•ืคื•ืจืฆื™ื•ื ืœื™ ื‘ื”ืชืงืคื” ืฉืœ ืฆื‘ื ืคืงื™ืกื˜ืŸ ื ื’ื“ ื”ืื•ื›ืœื•ืกื™ื™ื” ื”ื‘ื ื’ืœื™ืช ืฉืœ ืžื–ืจื— ืคืงื™ืกื˜ืŸ. ื‘ืžืืžืจ ื‘ืžื’ื–ื™ืŸ "ื˜ื™ื™ื" ืžื™ื•ื 2 ื‘ืื•ื’ื•ืกื˜ 1971 ื ืืžืจ "ื”ื”ื™ื ื“ื™ื, ื”ืžื”ื•ื•ื™ื ืฉืœื•ืฉื” ืจื‘ืขื™ื ืžื”ืคืœื™ื˜ื™ื ื•ืจื•ื‘ ื”ื”ืจื•ื’ื™ื, ื”ื™ื• ืขื™ืงืจ ื”ื ืคื’ืขื™ื ืขืœ ื™ื“ื™ ื”ืฉื ืื” ื”ืฆื‘ืื™ืช ื”ืžื•ืกืœืžื™ืช." ื”ืกื ืื˜ื•ืจ ืื“ื•ืืจื“ ืงื ื“ื™ ื›ืชื‘ ื‘ื“ื•"ื— ืž-1 ื‘ื ื•ื‘ืžื‘ืจ 1971, ืžื˜ืขื ื•ืขื“ืช ื”ืกื ืื˜ ืฉืœ ืืจืฆื•ืช ื”ื‘ืจื™ืช ืœื™ื—ืกื™ ื—ื•ืฅ, "ืืช ื”ืžื›ื•ืช ื”ืงืฉื•ืช ื‘ื™ื•ืชืจ ืกืคื’ื• ื—ื‘ืจื™ ื”ืงื”ื™ืœื” ื”ื”ื™ื ื“ื™ืช ืฉื ืฉื“ื“ื• ืžื”ื ืื“ืžื•ืชื™ื”ื ื•ื—ื ื•ื™ื•ืชื™ื”ื, ื ืฉื—ื˜ื• ื‘ืื•ืคืŸ ืฉื™ื˜ืชื™, ืขื‘ืจื• ืื•ื ืก ื”ืžื•ื ื™ ื‘ืžืงื•ืžื•ืช ืžืกื•ื™ืžื™ื, ื•ืกื•ืžื ื• ื‘ื˜ืœืื™ื ืฆื”ื•ื‘ื™ื ื”ืžืกื•ืžื ื™ื ื‘ืื•ืช 'H'. ื›ืœ ืืœื” ืื•ืฉืจื• ื‘ืื•ืคืŸ ืจืฉืžื™, ื”ื•ื–ืžื ื• ืžืืกืœืืžืื‘ืื“ ื•ื”ื•ืคืขืœื• ืขืœ ืคื™ ื—ื•ืง ืฆื‘ืื™". ืงื ื“ื™ ื“ื™ื•ื•ื— ื›ื™ 80% ืžื”ืคืœื™ื˜ื™ื ื‘ื”ื•ื“ื• ื”ื™ื• ื”ื™ื ื“ื™ื ื•ืœืคื™ ืกื•ื›ื ื•ื™ื•ืช ืกื™ื•ืข ื‘ื™ื ืœืื•ืžื™ื•ืช ื›ืžื• ืื•ื ืกืง"ื• ื•ืืจื’ื•ืŸ ื”ื‘ืจื™ืื•ืช ื”ืขื•ืœืžื™ ืžืกืคืจ ื”ืคืœื™ื˜ื™ื ืžืžื–ืจื— ืคืงื™ืกื˜ืŸ ื‘ืฉื™ื ืžืกืคืจื ื‘ื”ื•ื“ื• ื”ื™ื” ืงืจื•ื‘ ืœ-10 ืžื™ืœื™ื•ืŸ. ื‘ื˜ื•ืจ "ื”ืฉื—ื™ื˜ื” ื”ืคืงื™ืกื˜ื ื™ืช ืฉื ื™ืงืกื•ืŸ ื”ืชืขืœื ืžืžื ื”", ื›ืชื‘ ื”ืขื™ืชื•ื ืื™ ื–ื•ื›ื” ืคืจืก ืคื•ืœื™ืฆืจ, ืกื™ื“ื ื™ ืฉื ื‘ืจื’, ืขืœ ื—ื–ืจืชื• ืœื‘ื ื’ืœื“ืฉ ื”ืžืฉื•ื—ืจืจืช ื‘ืฉื ืช 1972. "ืชื–ื›ื•ืจื•ืช ืื—ืจื•ืช ื”ื™ื• ื”ืื•ืชื™ื•ืช 'H' ื”ืฆื”ื•ื‘ื•ืช ืฉืฆื™ื™ืจื• ื”ืคืงื™ืกื˜ื ื™ื ืขืœ ื‘ืชื™ื”ื ืฉืœ ื”ื”ื™ื ื“ื™ื, ื™ืขื“ื™ื ืžืกื•ื™ืžื™ื ืฉืœ ื”ืฆื‘ื ื”ืžื•ืกืœืžื™ (ื›ืœื•ืžืจ ืฆื‘ื ืคืงื™ืกื˜ืŸ, ืฉืคืขืœ ื’ื ื ื’ื“ ืžื•ืกืœืžื™ื ื‘ื ื’ืืœื™ื™ื)." (ื ื™ื•ื–ื“ื™ื™, 29 ื‘ืืคืจื™ืœ 1994). ื‘-28 ื‘ืคื‘ืจื•ืืจ 2013 ื“ืŸ ื‘ื™ืช ื”ื“ื™ืŸ ืœืคืฉืขื™ ืžืœื—ืžื” ืืช ืกื’ืŸ ื”ื ืฉื™ื ื“ืœื•ื•ืืจ ื—ื•ืกื™ื™ืŸ ืกืขื™ื“ื™ ืœืžื•ื•ืช ืขืœ ืคืฉืขื™ ืžืœื—ืžื” ืฉื‘ื•ืฆืขื• ื‘ืžื”ืœืš 1971 ื‘ืžืœื—ืžืช ื”ืขืฆืžืื•ืช ืฉืœ ื‘ื ื’ืœื“ืฉ. ืœืื—ืจ ื’ื–ืจ ื”ื“ื™ืŸ, ืคืขื™ืœื™ ืชื ื•ืขืช ื’'ืžืขืช-ืื™-ืื™ืกืœืืžื™ ื•ืื’ืฃ ื”ืกื˜ื•ื“ื ื˜ื™ื ืฉืœื” ืชืงืคื• ื”ื™ื ื“ื™ื ื‘ืื–ื•ืจื™ื ืฉื•ื ื™ื ื‘ืžื“ื™ื ื”. ื ื›ืกื™ื ื”ื™ื ื“ื™ื ื ื‘ื–ื–ื•, ื‘ืชื™ื ื”ื™ื ื“ื™ื ื ืฉืจืคื• ืœืืคืจ ื•ืžืงื“ืฉื™ื ื”ื™ื ื“ื™ื ื—ื•ืœืœื• ื•ื”ื•ืฆืชื•. ื”ืืœื™ืžื•ืช ื›ืœืœื” ื‘ื™ื–ืช ื ื›ืกื™ื ื•ืขืกืงื™ื ื”ื™ื ื“ื™ื, ืฉืจื™ืคืช ื‘ืชื™ื ื”ื™ื ื“ื™ื, ืื•ื ืก ื ืฉื™ื ื”ื™ื ื“ื™ื•ืช,[ื“ืจื•ืฉ ืžืงื•ืจ] ื—ื™ืœื•ืœ ืžื‘ื ื™ ืงื•ื“ืฉ ื•ื”ืจืก. ืœื“ื‘ืจื™ ืžื ื”ื™ื’ื™ ื”ืงื”ื™ืœื” ื™ื•ืชืจ ืž-50 ืžืงื“ืฉื™ื ื”ื™ื ื“ื™ื ื•-1,500 ื‘ืชื™ื ื”ื™ื ื“ื™ื ื ื”ืจืกื• ื‘-20 ืžื—ื•ื–ื•ืช. ื”ืžืžืฉืœื” ื˜ืขื ื” ืฉื”ืื—ืจื™ื•ืช ืขืœ ื”ืชืงืคื•ืช ืขืœ ื”ืžื™ืขื•ื˜ื™ื ื”ื™ื ืฉืœ ื’'ืžืขืช-ืื™-ืื™ืกืœืืžื™, ืืš ื”ื ื”ื’ืช ื’'ืžืขืช-ืื™-ืืกืœืืžื™ ื”ื›ื—ื™ืฉื” ื›ืœ ืžืขื•ืจื‘ื•ืช. ืจืืฉื™ ื”ืžื™ืขื•ื˜ื™ื ืžื—ื• ืขืœ ื”ื”ืชืงืคื•ืช. ื‘ื™ืช ื”ืžืฉืคื˜ ื”ืขืœื™ื•ืŸ ืฉืœ ื‘ื ื’ืœื“ืฉ ื”ื•ืจื” ืœืจืฉื•ื™ื•ืช ื”ื—ื•ืง ืœื”ืชื—ื™ืœ ื‘ื—ืงื™ืจืช ื”ืคื™ื’ื•ืขื™ื. ืฉื’ืจื™ืจ ืืจืฆื•ืช ื”ื‘ืจื™ืช ื‘ื‘ื ื’ืœื“ืฉ ื”ื‘ื™ืข ื“ืื’ื” ืžื”ืชืงืคืช ื’'ืžืขืช ืขืœ ื”ืงื”ื™ืœื” ื”ื”ื™ื ื“ื™ืช ื”ื‘ื ื’ืœื™ืช. ืจื“ื™ืคืช ื™ื”ื•ื“ื™ื ืื• ืื ื˜ื™ืฉืžื™ื•ืช ื”ื™ื ืชื•ืคืขื” ื—ื•ื–ืจืช ืœืื•ืจืš ื”ื”ื™ืกื˜ื•ืจื™ื” ื”ื™ื”ื•ื“ื™ืช. ื‘ืžืฉืš ื›ืžื” ืžืื•ืช ืฉื ื™ื ืจื“ื™ืคืช ื™ื”ื•ื“ื™ื ื”ื™ื™ืชื” ื‘ืขื™ืงืจ ืขืœ ืจืงืข ื“ืชื™, ื›ืš ืฉื™ื”ื•ื“ื™ื ื ืจื“ืคื• ืขืœ ืจืงืข ื“ืขื•ืช ืงื“ื•ืžื•ืช ืขืœ ืจืงืข ื”ืืžื•ื ื•ืช ื”ื“ืชื™ื•ืช ืฉืœื”ืŸ ืฉื”ื™ื• ืฉื•ื ื•ืช ืžืงื‘ื•ืฆื•ืช ื”ืจื•ื‘ โ€“ ื‘ืขื™ืงืจ ื ืฆืจื•ืช ื•ืืกืœืื. ื”ื“ื‘ืจ ื›ืœืœ ืงื™ื•ื ืฉืœ ืขืœื™ืœื•ืช ื“ื, ื—ื•ืงื™ื ืžืคืœื™ื ื ื’ื“ ื™ื”ื•ื“ื™ื, ื‘ื™ื–ื”, ืžืืกืจื™ื, ืขื™ื ื•ื™ื™ื, ืžืขืฉื™ ื”ืจื’ ืื• ืจืฆื— ืœืœื ืขื•ื ืฉ, ื—ื˜ื™ืคืช ื™ืœื“ื™ื ื•ื”ืžืจืช ื“ืชื, ืคื•ื’ืจื•ืžื™ื, ื•ื›ืŸ ื˜ื™ื”ื•ืจ ืืชื ื™ ืื• ื’ืจื•ืฉื™ื ื”ืžื•ื ื™ื™ื (ื›ืžื• ื’ื™ืจื•ืฉ ืื ื’ืœื™ื”, ื’ื™ืจื•ืฉ ืกืคืจื“. ืขื ื–ืืช ื™ื”ื•ื“ื™ื ืฉื”ืžื™ืจื• ืืช ื“ืชื ื”ื™ื• ื‘ืจื•ื‘ ื”ืžืงืจื™ื ื‘ื˜ื•ื—ื™ื. ื‘ืžื”ืœืš ื”ืžืื” ื”-19 ื”ื—ืœื” ื”ืื ื˜ื™ืฉืžื™ื•ืช ืœื”ืฉืชื ื•ืช ื•ื”ื›ื™ืœื” ื’ื ืžืžื“ ื’ื–ืขื ื™ โ€“ ืฉื‘ื• ื”ื™ื”ื•ื“ื™ื ื ื—ืฉื‘ื™ื ื›ืžื•ืฉื ืœืฉื ืื”, ื”ืคืœื™ื” ืจื“ื™ืคื” ื•ื”ืฉืžื“ื” ืขืœ ืจืงืข ืืชื ื™ ื•ืืฃ ื’ื ื˜ื™, ื›ืš ืฉื’ื ืฆืืฆืื™ื ืœื™ื”ื•ื“ื™ื ืฉืื™ื ื ื‘ื”ื›ืจื— ืžืงื™ื™ืžื™ื ืื•ืจื— ื—ื™ื™ื ื“ืชื™ ืื• ื›ืืœื” ืฉื”ืฉืชื™ื™ื›ื• ืœื“ืชื•ืช ืื—ืจื•ืช ื”ื™ื• ื ืชื•ื ื™ื ื‘ืกื›ื ื”. (ืจืื• ืคืจื˜ื™ื ื‘ื”ืžืฉืš). ืจื“ื™ืคืช ื”ืžื•ืกืœืžื™ื ื”ื™ื™ืชื” ืชื•ืคืขื” ื—ื•ื–ืจืช ืœืื•ืจืš ื”ื”ื™ืกื˜ื•ืจื™ื” ืฉืœ ื”ืืกืœืื: ืžืขืฆืจ ืœื ืžื•ืฆื“ืง, ืžืืกืจ, ืžื›ื•ืช, ืขื™ื ื•ื™ื™ื ืื• ื”ื•ืฆืื” ืœื”ื•ืจื’ ื•ื›ืŸ ื”ื—ืจืžื”, ื”ืฉืžื“ืช ืจื›ื•ืฉ, ืื• ื”ืกืชื” ืœืฉื ืืช ืžื•ืกืœืžื™ื. ืœืขื™ืชื™ื ื”ืชืจื—ื‘ื” ื”ืจื“ื™ืคื” ืžืขื‘ืจ ืœืืœื” ืฉืชืคืกื• ืืช ืขืฆืžื ื›ืžื•ืกืœืžื™ื ื•ื›ืœืœื” ืืช ืžื™ ืฉื ืชืคืก ื‘ืขื™ื ื™ ืื—ืจื™ื ื›ืžื•ืกืœืžื™ื, ืื• ืžื•ืกืœืžื™ื ืฉื ื—ืฉื‘ื• ืœื ืžื•ืกืœืžื™ื ื‘ืขื™ื ื™ ื”ืžื•ืกืœืžื™ื ื”ืื—ืจื™ื. ืื—ืžื“ื™ื ื”ื ืžื•ืกืœืžื™ื ื‘ื™ืขื ื™ ืขืฆืžื, ืืš ื‘ืขื™ื ื™ ืžื•ืกืœืžื™ื ืจื‘ื™ื ืื—ืจื™ื ื”ื ืœื ืžื•ืกืœืžื™ื ืืœื "ื›ื•ืคืจื™ื". ื‘ืฉื ืช 1984, ืžืžืฉืœืช ืคืงื™ืกื˜ืŸ, ื‘ืคื™ืงื•ื“ื• ืฉืœ ื”ื’ื ืจืœ ื–ื™ื-ืืœ-ื—ืืง, ื”ืขื‘ื™ืจื” ืืช ืคืงื•ื“ื” XX, ืืฉืจ ืืกืจื” ืขืœ ืื—ืžื“ื™ื ืœื”ืžื™ืจ ื“ืชื ืฉืœ ืื—ืจื™ื ืœื“ืชื-ืฉืœื”ื ื•ื›ืŸ ืืกืจื” ืขืœ ืื—ืžื“ื™ื ืœื”ืชื™ื™ื—ืก ืœืขืฆืžื ื›ืžื•ืกืœืžื™ื. ืขืœ ืคื™ ืคืงื•ื“ื” ื–ื•, ื›ืœ ืื—ืžื“ื™ ื”ืžืชื™ื™ื—ืก ืœืขืฆืžื• ื›ืžื•ืกืœืžื™ ื‘ื›ืชื‘ ืื• ืขืœ ื™ื“ื™ ื‘ื™ื˜ื•ื™ ื’ืœื•ื™, ื‘ืžื™ืฉืจื™ืŸ ืื• ื‘ืขืงื™ืคื™ืŸ, ืื• ืขื•ืฉื” ืืช ื”ืงืจื™ืื” ืœืชืคื™ืœื” ื›ืคื™ ืฉืขื•ืฉื™ื ืžื•ืกืœืžื™ื ืื—ืจื™ื, ื“ื™ื ื• ืžืืกืจ ืฉืœ ืขื“ 3 ืฉื ื™ื. ื‘ื’ืœืœ ื”ืงืฉื™ื™ื ื”ืœืœื•, ืžื™ืจื–ื” ื˜ื”ื™ืจ ืื—ืžื“ ื ื“ื“ ืœืœื•ื ื“ื•ืŸ.[ื“ืจื•ืฉ ืžืงื•ืจ] ื”ืคืจืขื•ืช ื”ืื ื˜ื™-ืกื™ืงื™ื•ืช ื‘-1984 (ื˜ื‘ื— ื”ืกื™ืงื™ื) ื”ื™ื• ืกื“ืจืช ืคื•ื’ืจื•ืžื™ื ื ื’ื“ ืกื™ืงื™ื ื‘ื”ื•ื“ื•, ืขืœ ื™ื“ื™ ืืกืคืกื•ืฃ ืื ื˜ื™-ืกื™ืงื™, ื‘ืชื’ื•ื‘ื” ืœื”ืชื ืงืฉื•ืช ื‘ื—ื™ื™ื” ืฉืœ ืื™ื ื“ื™ืจื” ื’ื ื“ื™, ื‘-31 ื‘ืื•ืงื˜ื•ื‘ืจ 1984, ืขืœ ื™ื“ื™ ืฉื ื™ื™ื ืžืฉื•ืžืจื™ ืจืืฉื” ื”ืกื™ืงื™ื ื‘ืชื’ื•ื‘ื” ืœืžืขืฉื™ื” ื”ืžืืฉืจื™ื ืืช ื”ืžื‘ืฆืข ื”ืฆื‘ืื™ ืžื‘ืฆืข ื”ื›ื•ื›ื‘ ื”ื›ื—ื•ืœ. ื™ื•ืชืจ ืž-8,000 ืกื™ืงื™ื ื ื”ืจื’ื•, ื›ื•ืœืœ 3,000 ื‘ื“ืœื”ื™. ื‘ื™ื•ื ื™ 1984, ื‘ืžื”ืœืš ืžื‘ืฆืข ื”ื›ื•ื›ื‘ ื”ื›ื—ื•ืœ, ื”ื•ืจืชื” ืื™ื ื“ื™ืจื” ื’ื ื“ื™ ืœืฆื‘ื ื”ื•ื“ื• ืœืชืงื•ืฃ ืืช ืžืงื“ืฉ ื”ื–ื”ื‘ ื•ืœื—ืกืœ ืืช ื”ืžืชืงื•ืžืžื™ื, ืกื™ืงื™ื ื‘ื“ืœื ื™ื ืฉืื’ืจื• ื ืฉืง. ืžืื•ื—ืจ ื™ื•ืชืจ ื”ื•ืคืขืœื• ื›ื•ื—ื•ืช ืฆื‘ืื™ื™ื ื”ื•ื“ื™ื ืœืคื™ื ื•ื™ ื”ื‘ื“ืœื ื™ื ืžื”ืื–ื•ืจ ื”ื›ืคืจื™ ืฉืœ ืžื“ื™ื ืช ืคื ื’'ืื‘. ืžืžืฉืœืช ื”ื•ื“ื• ื“ื™ื•ื•ื—ื” ืขืœ 2,700 ื”ืจื•ื’ื™ื ื‘ืชื•ื”ื• ื•ื‘ื•ื”ื• ืฉืœืื—ืจ ืžื›ืŸ. ื‘ืขืงื‘ื•ืช ื”ืžื”ื•ืžื•ืช ื“ื™ื•ื•ื—ื” ืžืžืฉืœืช ื”ื•ื“ื• ื›ื™ 20,000 ื‘ืจื—ื• ืžื”ืขื™ืจ. ื”ืื™ื—ื•ื“ ื”ืขืžืžื™ ืœื—ื™ืจื•ื™ื•ืช ืื–ืจื—ื™ื•ืช ื“ื™ื•ื•ื— ืขืœ ืœืคื—ื•ืช 1,000 ืขืงื•ืจื™ื. ื”ืื–ื•ืจื™ื ื”ืžื•ืฉืคืขื™ื ื‘ื™ื•ืชืจ ื”ื™ื• ืฉื›ื•ื ื•ืช ื”ืกื™ืงื™ื ื‘ื“ืœื”ื™. ื”ืœืฉื›ื” ื”ืžืจื›ื–ื™ืช ืœื—ืงื™ืจื”, ืกื•ื›ื ื•ืช ื”ื—ืงื™ืจื•ืช ื”ื”ื•ื“ื™ืช ื”ืจืืฉื™ืช, ืกื‘ืจื” ื›ื™ ืžืขืฉื™ ื”ืืœื™ืžื•ืช ื”ืชืงื™ื™ืžื• ื‘ืชืžื™ื›ืชื ืฉืœ ืื ืฉื™ ืžืฉื˜ืจืช ื“ืœื”ื™ ื“ืื– ื•ื”ืžืžืฉืœื” ื”ืžืจื›ื–ื™ืช ื‘ืจืืฉื•ืช ื‘ื ื” ืฉืœ ืื™ื ื“ื™ืจื” ื’ื ื“ื™, ืจื’'ื™ื‘ ื’ื ื“ื™. ืจื’'ื™ื‘ ื’ื ื“ื™ ื”ื•ืฉื‘ืข ืœื›ื”ืŸ ื›ืจืืฉ ืžืžืฉืœื” ืœืื—ืจ ืžื•ืช ืืžื•, ื•ื›ืฉื ืฉืืœ ืขืœ ื”ืคืจื•ืช ืกื“ืจ, ืืžืจ "ื›ืืฉืจ ืขืฅ ื’ื“ื•ืœ ื ื•ืคืœ ื”ืื“ืžื” ืจื•ืขื“ืช" ื•ื‘ื›ืš ื ื™ืกื” ืœื”ืฆื“ื™ืง ืืช ื”ืกื›ืกื•ืš ื”ืงื”ื™ืœืชื™. ื™ืฉ ื˜ืขื ื•ืช ืฉื”ืžืžืฉืœื” ื”ืฉืžื™ื“ื” ืจืื™ื•ืช ื•ื”ื’ื ื” ืขืœ ื”ืืฉืžื™ื ื‘ืจื“ื™ืคื”. ืขื™ืชื•ืŸ ืžืจื›ื–ื™ ื›ื™ื ื” ืืช ืคืขื•ืœื•ืช ื”ืžืžืฉืœื” "ืื ื›ืœ ื”ื˜ื™ื•ื—ื™ื". ื ื˜ืขืŸ ืฉื”ืืœื™ืžื•ืช ื”ื•ื‘ืœื” ื•ืœืขื™ืชื™ื ืงืจื•ื‘ื•ืช ื‘ื•ืฆืขื” ืขืœ ื™ื“ื™ ืคืขื™ืœื™ ื”ืงื•ื ื’ืจืก ื”ืœืื•ืžื™ ื”ื”ื•ื“ื™ ื•ืื•ื”ื“ื™ื ื‘ืžื”ืœืš ื”ื”ืชืคืจืขื•ื™ื•ืช. ื”ื ืฉืง ื”ืจืืฉื™ ืฉืฉื™ืžืฉ ืืช ื”ื”ืžื•ื ื™ื, ื ืคื˜, ืกื•ืคืง ืขืœ ื™ื“ื™ ืงื‘ื•ืฆื” ืฉืœ ืžื ื”ื™ื’ื™ ืžืคืœื’ืช ื”ืงื•ื ื’ืจืก ื”ืœืื•ืžื™ ื”ื”ื•ื“ื™, ืฉื”ื™ื• ื‘ื‘ืขืœื•ืช ืชื—ื ื•ืช ื“ืœืง. ืจื“ื™ืคื•ืช ืืชื ื™ื•ืช ืจื“ื™ืคื” ืืชื ื™ืช ื”ื™ื ืจื“ื™ืคื” ื”ืžื‘ื•ืกืกืช ืขืœ ืžื•ืฆื ืืชื ื™. ื–ื•ื”ื™ ืจื“ื™ืคื” ื“ื•ืžื” ืœืจื“ื™ืคื” ืขืœ ืจืงืข ื’ื–ืขื ื™ (ืขืœ ื‘ืกื™ืก ื’ื–ืข). ืจืฆื— ื”ืขื ื‘ืจื•ืื ื“ื” ืฉืœ ื”ื˜ื•ื˜ืกื™ ืขืœ ื™ื“ื™ ื”ื•ื˜ื• ื ืžืฆื ื‘ืงื˜ื’ื•ืจื™ื” ื–ื•. ื”ื›ื™ื‘ื•ืฉ ื”ื™ืคื ื™ ื‘ืกื™ืŸ ื’ืจื ืœืžื•ืชื ืฉืœ ืžื™ืœื™ื•ื ื™ ืื ืฉื™ื, ื‘ืขื™ืงืจ ืื™ื›ืจื™ื ืฉื ืจืฆื—ื• ืœืื—ืจ ืคืฉื™ื˜ืช ื“ื•ืœื™ื˜ืœ ื‘ืชื—ื™ืœืช ืžืœื—ืžืช ื”ืขื•ืœื ื”ืฉื ื™ื™ื”. ื‘ืžืื” ื”-19 ื”ืื ื˜ื™ืฉืžื™ื•ืช ื”ื—ืœื” ืœืฉื ื•ืช ืืช ืื•ืคื™ื™ื” โ€“ ืœื ืจืง ืจื“ื™ืคื” ืฉืœ ื™ื”ื•ื“ื™ื ื‘ื’ืœืœ ื“ืชื ื”ืฉื•ื ื”, ืืœื” ืจื“ื™ืคื” ืฉืœ ื™ื”ื•ื“ื™ื ื’ื ืขืœ ืจืงืข ื”ืžื•ืฆื ื”ืืชื ื™ ืฉืœื”ื. ืชืื•ืจื™ื•ืช ืื ื˜ื™ืฉืžื™ื•ืช ืžื•ื“ืจื ื™ื•ืช ื”ื’ื™ืขื• ื•ื‘ื”ืŸ ืชื•ืจืช ื”ื’ื–ืข. ืฉื™ืื” ืฉืœ ืจื“ื™ืคื” ื–ื• ื”ื™ื™ืชื” ื”ืฉื•ืื” ื‘ืžืœื—ืžืช ื”ืขื•ืœื ื”ืฉื ื™ื™ื” ื‘ื” ื ืจืฆื—ื• ื›-6 ืžื™ืœื™ื•ืŸ ืื™ืฉ ืจืง ื‘ื’ืœืœ ื”ื™ื•ืชื ื™ื”ื•ื“ื™ื ืื• ืฆืืฆืื™ื ืฉืœ ื™ื”ื•ื“ื™ื. ื‘ืžืื” ื”-20 ื•ืื™ืœืš ืื ื˜ื™ืฉืžื™ื•ืช ื‘ืื” ืœื™ื“ื™ ื‘ื™ื˜ื•ื™ ืœืขื™ืชื™ื ืงืจื•ื‘ื•ืช ื›ืื ื˜ื™-ืฆื™ื•ื ื•ืช, ืœืžืจื•ืช ื”ืขื•ื‘ื“ื” ืฉื™ืฉื ืŸ ืงื‘ื•ืฆื•ืช ื™ื”ื•ื“ื™ื•ืช ืฉื•ื ื•ืช ืฉื—ื‘ืจื™ื”ืŸ ืžืชื ื’ื“ื™ื ืœืจืขื™ื•ืŸ ื”ืฆื™ื•ื ื•ืช. ื›ืžื• ื›ืŸ ื™ื”ื•ื“ื™ื ื ืจื“ืคื™ื ื”ืŸ ื‘ืฉืœ ื”ืžื•ืฆื ื”ืืชื ื™ ืฉืœื”ื ื•ื”ืŸ ื‘ืฉืœ ื“ืชื. ื‘ืื™ืžืคืจื™ื” ื”ืขื•ืช'ืžืื ื™ืช ื—ื™ื• ื‘ืืžืฆืข ื”ืžืื” ื”-19 ื›-2.5 ืžื™ืœื™ื•ืŸ ืืจืžื ื™ื. ืขืœ ืจืงืข ืงืฉื™ื™ื ืฉืœ ื”ืื™ืžืคืจื™ื” ื•ืขืœื™ื” ื‘ืžื•ื“ืขื•ืช ืœืœืื•ืžื™ื•ืช ื”ื—ืœื• ืจื“ื™ืคื•ืช ื ื’ื“ ืืจืžื ื™ื ืฉืœื”ื›ื• ื•ื”ื—ืจื™ืคื•. ื‘ืฉืœื”ื™ ื”ืžืื” ื”ืžืขืฆืžื•ืช ื”ื’ื“ื•ืœื•ืช ื‘ืชืงื•ืคื” ื“ื ื• ื‘ื ื•ืฉื "ื”ืฉืืœื” ื”ืืจืžื ื™ืช" ืžื” ื™ืฉ ืœืขืฉื•ืช ื›ื“ื™ ืœื”ื’ืŸ ืขืœ ื”ืืจืžื ื™ื. ื”ืจื“ื™ืคื•ืช ื ื’ื“ ื”ืืจืžื ื™ื ื’ืจืžื• ืœืžืกืคืจ ืžืขืฉื™ ื˜ื‘ื— ืจื—ื‘ื™ ื”ื™ืงืฃ ื›ืžื• ืžืขืฉื™ ื”ื˜ื‘ื— ื”ื—ืžื™ื“ื™ื™ื ื•ื˜ื‘ื— ืื“ื ื”. ืœื‘ืกื•ืฃ ืจื•ื‘ ื”ืืจืžื ื™ื ื ืจืฆื—ื• ื•ื”ื™ืชืจ ื”ื•ื’ืœื• ื‘ืžืกื’ืจืช ืจืฆื— ื”ืขื ื”ืืจืžื ื™. ื‘ืžืงื‘ื™ืœ ื ืจืฆื—ื• ื•ื’ื•ืจืฉื• ืžื™ืขื•ื˜ื™ื ื ื•ืฆืจื™ื ืื—ืจื™ื ื‘ืžืกื’ืจืช ืจืฆื— ื”ืขื ื”ืืฉื•ืจื™ ื•ืจืฆื— ื”ืขื ื”ื™ื•ื•ื ื™-ืคื•ื ื˜ื™. ืจื“ื™ืคืช ื’ืจืžื ื™ื ืืชื ื™ื™ื ืžืชื™ื™ื—ืกืช ืœืคืขื™ืœื•ืช ืฉื™ื˜ืชื™ืช ื ื’ื“ ืงื‘ื•ืฆื•ืช ืฉืœ ื’ืจืžื ื™ื ืืชื ื™ื™ื ืขืœ ืกืžืš ืžื•ืฆืื ื”ืืชื ื™. ืžื‘ื—ื™ื ื” ื”ื™ืกื˜ื•ืจื™ืช, ืจื“ื™ืคื” ื–ื• ื ื‘ืขื” ืžืฉืชื™ ืกื™ื‘ื•ืช: ื”ืื•ื›ืœื•ืกื™ื™ื” ื”ื’ืจืžื ื™ืช ื ื—ืฉื‘ื”, ื‘ื™ืŸ ืื ื‘ืฆื“ืง ืื• ืœื, ืงืฉื•ืจื” ืœืžืฉื˜ืจื™ื ืœืื•ืžื ื™ื™ื ื’ืจืžื ื™ื ื›ืžื• ื”ื ืืฆื™ื–ื ืื• ื”ืงื™ืกืจ ื•ื™ืœื”ืœื. ื‘ืชืงื•ืคืช ืžืœื—ืžืช ื”ืขื•ืœื ื”ืจืืฉื•ื ื” ื ืจื“ืคื• ื’ืจืžื ื™ื ื‘ืืจืฆื•ืช ื”ื‘ืจื™ืช, ื›ืš ืงืจื” ื’ื ื‘ืžื–ืจื— ื•ืžืจื›ื– ืื™ืจื•ืคื” ืœืื—ืจ ืกื™ื•ื ืžืœื—ืžืช ื”ืขื•ืœื ื”ืฉื ื™ื™ื”. ืœืจื•ื‘ ื”ืงื•ืจื‘ื ื•ืช ืฉืœ ืจื“ื™ืคื•ืช ืืœื” ืœื ื”ื™ื” ืงืฉืจ ืœืื•ืชื ืžืฉื˜ืจื™ื, ืืš ื”ื™ื” ืฉื™ืชื•ืฃ ืคืขื•ืœื” ื‘ื™ืŸ ืืจื’ื•ื ื™ ื”ืžื™ืขื•ื˜ื™ื ื”ื’ืจืžื ื™ื ื•ื”ืžืฉื˜ืจ ื”ื ืืฆื™, ื›ืคื™ ืฉืžืจืื” ื”ื“ื•ื’ืžื” ืฉืœ ื–ืœื‘ืกื˜ืฉื•ืฅ (ืื '), ื”ืžืฉืžืฉ ืขื“ื™ื™ืŸ ื›ืชื™ืจื•ืฅ ืœืคืขื•ืœื•ืช ืื™ื‘ื” ื’ื ื ื’ื“ ืืœื” ืฉืœื ื”ืฉืชืชืคื• ื‘ืืจื’ื•ื ื™ื ืืœื”. ืœืื—ืจ ืžืœื—ืžืช ื”ืขื•ืœื ื”ืฉื ื™ื™ื”, ืจื‘ื™ื ืžืคื•ืœืงืกื“ื•ื™ื˜ืฉื” ืืœื” ื ื”ืจื’ื• ืื• ื’ื•ืจืฉื• ืžื‘ืชื™ื”ื ื‘ืžืขืฉื™ ื ืงื, ื•ืื—ืจื™ื โ€“ ื‘ื˜ื™ื”ื•ืจ ืืชื ื™ ืฉืœ ืฉื˜ื—ื™ื ืœืคื ื™ ืื›ืœื•ืกื ื‘ืื–ืจื—ื™ ื”ืžื“ื™ื ื” ื”ืžืกืคื—ืช. ื‘ืžืงืจื™ื ืื—ืจื™ื (ืœืžืฉืœ ื‘ืžืงืจื” ืฉืœ ื”ืื•ื›ืœื•ืกื™ื•ืช ื”ื’ื“ื•ืœื•ืช ืœืฉืขื‘ืจ ืฉื”ื™ื• ื“ื•ื‘ืจื•ืช ื’ืจืžื ื™ืช ื‘ืจื•ืกื™ื” ืื• ืืกื˜ื•ื ื™ื”, ืื• ื”ืžื™ืขื•ื˜ ื”ื’ืจืžื ื™ ื”ื˜ืจื ืกื™ืœื•ื•ื ื™ (ื–ื™ื‘ื ื‘ื•ืจื’ืŸ) ื‘ืจื•ืžื ื™ื” ื•ื‘ื‘ืœืงืŸ), ื”ื™ื™ืชื” ืจื“ื™ืคื” ืฉืœ ืงื”ื™ืœื•ืช ื—ืคื•ืช ืžืคืฉืข ืฉืœื ื”ื™ื• ื—ืœืง ืžื”ืจื™ื™ืš ื”ืฉืœื™ืฉื™. ืื ืฉื™ ื”ื–ืืจื” ื‘ืžืจื›ื– ืืคื’ื ื™ืกื˜ืŸ ื ืจื“ืคื• ืขืœ ื™ื“ื™ ืฉืœื™ื˜ื™ ืืคื’ื ื™ืกื˜ืŸ ื‘ืชืงื•ืคื•ืช ืฉื•ื ื•ืช ื‘ื”ื™ืกื˜ื•ืจื™ื”. ืžืื– ืคื™ื’ื•ืขื™ 11 ื‘ืกืคื˜ืžื‘ืจ 2001, ืžื—ื‘ืœื™ื ืžื•ืกืœืžื™ื ืกื•ื ื™ื ืชื•ืงืคื™ื ืืช ืงื”ื™ืœืช ื”ื–ืืจื” ื‘ื“ืจื•ื ืžืขืจื‘ ื”ืขื™ืจ ื›ื•ื•ื™ืชื” ื‘ืคืงื™ืกื˜ืŸ, ื‘ื™ืชื ืฉืœ ื›-500,000 ื‘ื ื™ ื”ื–ืืจื” ืฉื ืžืœื˜ื• ืžืจื“ื™ืคื•ืช ื‘ืืคื’ื ื™ืกื˜ืŸ ื”ืฉื›ื ื”. ื›-2,400 ื’ื‘ืจื™ื, ื ืฉื™ื ื•ื™ืœื“ื™ื ื ื”ืจื’ื• ืื• ื ืคืฆืขื• ื›ืืฉืจ ืœืฉืงืจ-ืื”-ื’'ื ื’ื•ื•ื™ (ืื ') ืœืงื— ืื—ืจื™ื•ืช ืขืœ ืžืจื‘ื™ืช ื”ื”ืชืงืคื•ืช ื ื’ื“ ื”ืงื”ื™ืœื”. ื›ืชื•ืฆืื” ืžื›ืš, ืืœืคื™ื ืจื‘ื™ื ื‘ืจื—ื• ืžื”ืžื“ื™ื ื” ื•ื‘ื™ืงืฉื• ืžืงืœื˜ ื‘ืื•ืกื˜ืจืœื™ื”. ืื ื˜ื™-ืฆื•ืขื ื™ื•ืช ื”ื™ื ืขื•ื™ื ื•ืช, ื“ืขื•ืช ืงื“ื•ืžื•ืช, ืืคืœื™ื” ืื• ื’ื–ืขื ื•ืช ื”ืžื•ืคื ื™ืช ื›ืœืคื™ ื”ืฆื•ืขื ื™ื ื›ืงื‘ื•ืฆื” ืืชื ื™ืช, ืื• ืื ืฉื™ื ื”ื ืชืคืกื™ื ื›ื‘ื ื™ ื”ืžื•ืจืฉืช ื”ืฆื•ืขื ื™ืช. ื‘ืžื”ืœืš ืžืœื—ืžืช ื”ืขื•ืœื ื”ืฉื ื™ื™ื” ื ื™ืกื• ืžืžืฉืœืช ื’ืจืžื ื™ื” ื”ื ืืฆื™ืช ื•ื‘ืขืœื•ืช ื‘ืจื™ืชื” ืœื”ืฉืžื™ื“ ืืช ื”ืขื ื”ืฆื•ืขื ื™ ื‘ืื™ืจื•ืคื”, ื‘ืžืืžืฅ ืžืชื•ื›ื ืŸ ื•ื‘ื ื™ืกื™ื•ืŸ, ืฉืชื•ืืจ ืœืขื™ืชื™ื ืงืจื•ื‘ื•ืช ื›ื’'ื ื•ืกื™ื™ื“. ืชื—ืช ืฉืœื˜ื•ื ื• ืฉืœ ืื“ื•ืœืฃ ื”ื™ื˜ืœืจ, ื”ื•ืฆื ืฆื• ืžืฉืœื™ื ืœื—ื•ืงื™ ื ื™ืจื ื‘ืจื’ ื‘ึพ26 ื‘ื ื•ื‘ืžื‘ืจ 1935, ื•ื”ื’ื“ื™ืจ ืืช ื”ืฆื•ืขื ื™ื ื›"ืื•ื™ื‘ื™ ืžื“ื™ื ืช ื”ื’ื–ืข", ืื•ืชื” ืงื˜ื’ื•ืจื™ื” ื›ืžื• ื”ื™ื”ื•ื“ื™ื. ืœืคื™ื›ืš, ื’ื•ืจืœื ืฉืœ ื”ืฆื•ืขื ื™ื ื‘ืื™ืจื•ืคื” ืืžื•ืจ ื”ื™ื” ืœื”ืงื‘ื™ืœ ื‘ืžื•ื‘ืŸ ืžืกื•ื™ื ืœื–ื” ืฉืœ ื”ื™ื”ื•ื“ื™ื. ื”ื™ืกื˜ื•ืจื™ื•ื ื™ื ืžืขืจื™ื›ื™ื ื›ื™ 220,000 ืขื“ 500,000 ืฆื•ืขื ื™ื ื ืจืฆื—ื• ืขืœ ื™ื“ื™ ื”ื ืืฆื™ื ื•ืžืฉืชืคื™ ื”ืคืขื•ืœื” ืขืžื, ืื• ื™ื•ืชืจ ืž-25% ืžื›ืœืœ ืคื—ื•ืช ืžืžื™ืœื™ื•ืŸ ืฆื•ืขื ื™ื ืฉื—ื™ื• ื‘ืื™ืจื•ืคื” ื‘ืื•ืชื” ืชืงื•ืคื”. ืื™ืืŸ ื”ื ืงื•ืง ื˜ื•ืขืŸ ืฉืžืกืคืจ ื”ื”ืจื•ื’ื™ื ืžืงืจื‘ื ื”ื™ื” 1.5 ืžื™ืœื™ื•ืŸ. ืจืืฉ ื”ืžื—ืœืงื” ืœื–ื›ื•ื™ื•ืช ื”ืื“ื ืฉืœ ื”ืื•"ื ื’ื™ื ื” ืืช "ื”ืžืชืงืคื” ื”ืฉื™ื˜ืชื™ืช" ืœื›ืื•ืจื” ืฉืœ ืžื™ืื ืžืจ ืขืœ ืžื™ืขื•ื˜ ื”ืจื•ื”ื™ื ื’ื™ื”, ื•ื”ื–ื”ื™ืจ ื›ื™ ื ืจืื” ื›ื™ "ื˜ื™ื”ื•ืจ ืืชื ื™" ืžืชื ื”ืœ. ืžื•ืกืœืžื™ื ื‘ื ื™ ื”ืจื•ื”ื™ื ื’ื™ื” ืฉื ืžืœื˜ื• ืžื›ื•ื—ื•ืช ื”ื‘ื™ื˜ื—ื•ืŸ ื‘ืžื“ื™ื ืช ืจืื—ื™ืŸ ืฉืœ ืžื™ืื ืžืจ ืชื™ืืจื• ื”ืจื™ื’ื•ืช, ื”ืคื’ื–ื•ืช ื•ื”ืฆืชื” ื‘ื›ืคืจื™ื”ื, ื”ื›ื•ืœืœื™ื ืืช ื›ืœ ื”ืกืžืžื ื™ื ืฉืœ ืžืกืข "ื˜ื™ื”ื•ืจ ืืชื ื™", ืืžืจ ืืจื’ื•ืŸ Human Rights Watch. "ืœืคืœื™ื˜ื™ ืจื•ื”ื™ื ื’ื™ื” ื™ืฉ ืชื™ืื•ืจื™ื ืžื—ืจื™ื“ื™ื ืฉืœ ื‘ืจื™ื—ื” ืžื”ืชืงืคื•ืช ืฆื‘ื ื‘ื•ืจืžื–ื™ ื•ืฆืคื™ื™ื” ื‘ื›ืคืจื™ื”ื ื ื”ืจืกื™ื", ืืžืจ ืžื™ื ืืงืฉื™ ื’ืื ื’ื•ืœื™, ืžื ื”ืœ ื“ืจื•ื ืืกื™ื” ืฉืœ ื”ืืจื’ื•ืŸ. "ืคืขื•ืœื•ืช ื—ื•ืงื™ื•ืช ื ื’ื“ ืงื‘ื•ืฆื•ืช ื—ืžื•ืฉื•ืช ืื™ื ืŸ ื›ืจื•ื›ื•ืช ื‘ื”ื‘ืจื—ืช ื”ืื•ื›ืœื•ืกื™ื™ื” ื”ืžืงื•ืžื™ืช ืžื‘ืชื™ื”." ื”ืชืงืคื•ืช ื ืจื—ื‘ื•ืช ืขืœ ื˜ืžื™ืœื™ื ื‘ืกืจื™ ืœื ืงื” ืœื‘ืฉื• ืฆื•ืจื” ืฉืœ ืคืจืขื•ืช ืืชื ื™ื•ืช ืจื—ื‘ื•ืช ื‘ืื™, ื›ื•ืœืœ ื”ืคื•ื’ืจื•ื ื”ืื ื˜ื™ ื˜ืžื™ืœื™ ื‘ืฉื ืช 1958 ื•ืคืจืขื•ืช ื™ื•ืœื™ ื”ืฉื—ื•ืจ ืฉื›ืœืœื• ืจื“ื™ืคื•ืช ื ื•ืกืคื•ืช ื‘ืขืงื‘ื•ืช ืžืขืฉื™ ืจืฆื—, ืื•ื ืก ื•ื—ื˜ื™ืคื”. ืงื•ื“ื ืœื›ืŸ, ืจื•ื‘ ื”ื˜ืžื™ืœื™ื ื“ืจืฉื• ืœื”ืงื™ื ืžื“ื™ื ื” ื ืคืจื“ืช, ืืš ื‘-1983 ื”ื—ืœื• ืžืื‘ืงื™ื ืžื–ื•ื™ื ื™ื ื ื’ื“ ืงื™ืฆื•ื ื™ื ืกื™ื ื”ืœื™ื (ื”ืงื‘ื•ืฆื” ื”ืืชื ื™ืช ื”ื’ื“ื•ืœื” ื‘ื™ื•ืชืจ ื‘ืกืจื™ ืœื ืงื”), ืฉื”ื’ื™ืขื• ืœืฉื™ืื ื‘ื”ืงืžืช ืืจื’ื•ืŸ ื˜ื™ื’ืจื™ืกื™ ื”ืฉื—ืจื•ืจ ืฉืœ ื˜ืืžื™ืœ ืื™ืœื. ืจื“ื™ืคื” ื”ืžื‘ื•ืกืกืช ืขืœ ื‘ืกื™ืก ืฉื•ื ื™ ื’ื•ืคื ื™ ืื• ื ืคืฉื™ ื‘ืชืจื‘ื•ื™ื•ืช ืžืกื•ืจืชื™ื•ืช ืจื‘ื•ืช ืจืื• ื‘ื‘ืขืœื™ ืžื•ืžื™ื (ื ื›ื™ ื’ืคื™ื, ื—ืจืฉื™ื, ืขื™ื•ื•ืจื™ื), ืžืชืžื•ื“ื“ื™ ืžื—ืœื•ืช ื ืคืฉ ืื• ืžืคื’ืจื™ื ื›ืกื•ื’ ืฉืœ ืขื•ื ืฉ ืžืืœื•ื”ื™ื ืื• ืงืœืœื” ื‘ืฉืœ ื—ื˜ืื™ื ืฉื•ื ื™ื. ื‘ื—ื‘ืจื•ืช ืฉื•ื ื•ืช ื”ื™ื” ื ื”ื•ื’ ืœื”ื’ืœื•ืช ื—ืœืง ืžื”ืกื•ื‘ืœื™ื ืžืžื—ืœื•ืช ื›ื’ื•ืŸ ืžืฆื•ืจืขื™ื ื”ืŸ ื‘ื’ืœืœ ืคื—ื“ ืžื”ื“ื‘ืงื” ื•ื”ืŸ ื‘ื’ืœืœ ื“ืขื•ืช ืงื“ื•ืžื•ืช ืœืคื™ื”ืŸ ืžื—ืœื” ื ื•ื‘ืขืช ืžืขื•ื ืฉ ืืœื•ื”ื™ ื•ืœื›ืŸ ืื™ืŸ ื˜ืขื ืœืจื—ื ืขืœ ื”"ื—ื•ื˜ืื™ื". ื‘ืžื—ืฆื™ืช ื”ืจืืฉื•ื ื” ืฉืœ ื”ืžืื” ื”-20 ื ืขืฉื” ืฉื™ืžื•ืฉ ืžืกื™ื‘ื™ ื‘ืขื™ืงื•ืจ ื›ืคื•ื™ ื‘ืžื“ื™ื ื•ืช ืžืกื•ื™ืžื•ืช ื‘ืื™ืจื•ืคื” ื•ื‘ืฆืคื•ืŸ ืืžืจื™ืงื”, ื‘ื“ืจืš ื›ืœืœ ื›ื—ืœืง ืžืชื•ื›ื ื™ืช ืืื•ื’ื ื™ืช. ืชื•ืžื›ื™ ื”ืขื™ืงื•ืจ ื”ืฆื™ื’ื• ื–ืืช ื›ื ื™ืกื™ื•ืŸ ืœืคื˜ื•ืจ ืืช ื”ื—ื‘ืจื” ืžืžื—ืœื•ืช ื’ื ื˜ื™ื•ืช ื•ื‘ื”ืŸ ืžื•ื’ื‘ืœื•ืช ืฉื›ืœื™ืช ื”ืชืคืชื—ื•ืชื™ืช, ืžื—ืœื•ืช ื ืคืฉ ืžืกื•ื™ืžื•ืช ืื• ื”ืคืจืขื•ืช ื”ืชื ื”ื’ื•ืชื™ื•ืช ืžืกื•ื™ืžื•ืช. ื”ื™ืกื˜ื•ืจื™ื•ื ื™ื ืฉืœ ื”ืžื“ืข ื˜ื•ืขื ื™ื ืฉืžื—ืœื•ืช ื›ืืœื• ืฉื™ืฉ ืœื”ืŸ ืจืงืข ื’ื ื˜ื™ ื”ืŸ ื ื“ื™ืจื•ืช ื™ื—ืกื™ืช ื•ืขืœ ื›ืŸ ืขื™ืงื•ืจ ื”ืžื•ื ื™ ืื™ื ื• ื ื—ื•ืฅ ืœืฉื ืžื ื™ืขืชืŸ ื•ืื™ื ื• ืืคืงื˜ื™ื‘ื™, ื•ื”ืขื™ืงื•ืจ ืœื ื”ื™ื” ืืœื ื ื™ืกื™ื•ืŸ ืœืขืฆื‘ ืืช ื”ื—ื‘ืจื” ื‘ืื•ืคืŸ ืฉื™ืกืœืง ืžืžื ื” ื™ืกื•ื“ื•ืช ืฉื”ื•ื’ื“ืจื• ื›ื-ืกื•ืฆื™ืืœื™ื™ื, ื›ื“ื™ ืœืฉืžื•ืจ ืขืœ ืกื“ืจ ื—ื‘ืจืชื™ ืกืคืฆื™ืคื™ ืื• ืœืขืฆื‘ ืกื“ืจ ื—ื‘ืจืชื™ ืจืฆื•ื™. ืชื•ื›ื ื™ืช T4 - ืื•ืชื ืกื™ื” ื”ื™ื” ืžื‘ืฆืข ืฉื ืขืจืš ื‘ืžื”ืœืš ืฉืœื˜ื•ื ื ืฉืœ ื”ื ืืฆื™ื ื‘ื’ืจืžื ื™ื”, ื•ื ื•ืขื“ ืœื—ื™ืกื•ืœ ื›ืœ "ื”ืœื ื›ืฉื™ืจื™ื" ื•ืœื“ืื•ื’ ืœ"ืฉืžื™ืจืช ื˜ื•ื”ืจ ื”ื’ื–ืข ื”ืืจื™". ื‘ืชื•ื›ื ื™ืช ื–ื•, ืฉื”ืชื‘ืกืกื” ืขืœ ื”ืื™ื’ื ื™ืงื” ื”ื ืืฆื™ืช ื•ืขืœ ืขืงืจื•ื ื•ืช ื”ืฉื‘ื—ื” ื’ื–ืขื™ืช, ืจืฆื—ื• ื”ื ืืฆื™ื ื‘ืืžืฆืขื™ื ืฉื•ื ื™ื, ืœืจื‘ื•ืช ืชืื™ ื’ื–ื™ื, ื›-90 ืืœืฃ ื’ืจืžื ื™ื ืฉืกื‘ืœื• ืžืคื’ืžื™ื ื’ื•ืคื ื™ื™ื, ื›ืžื•ืžื™ื ืžื•ืœื“ื™ื ืื• ืžื—ืœื•ืช ื›ืจื•ื ื™ื•ืช ื•ื›ืŸ ืžืงืฉื™ื™ื ื ืคืฉื™ื™ื. ืฉืžื” ืฉืœ ื”ืชื•ื›ื ื™ืช, T4, ื‘ื ืžื›ืชื•ื‘ืช ืžื˜ื” ื”ืชื•ื›ื ื™ืช ื‘ืจื—' ื˜ื™ืจื’ืืจื˜ืŸ 4 ืฉื‘ื‘ืจืœื™ืŸ. ืจืขื™ื•ื ื•ืช ื•ืืžืฆืขื™ื ืฉื™ื•ืฉืžื• ื‘ืชื•ื›ื ื™ืช ื–ื•, ืฉื ื•ื”ืœื” ืขืœ ื™ื“ื™ ื”ื ืืฆื™ื, ื”ื•ืขืชืงื• ื‘ืฉืœื‘ ืžืื•ื—ืจ ื™ื•ืชืจ ืœืฉื•ืื”. ืจื“ื™ืคื” ืขืœ ื‘ืกื™ืก ืœื‘ืงื ื•ืช ืžื‘ื•ืกืกืช ืœืขื™ืชื™ื ืงืจื•ื‘ื•ืช ืขืœ ื”ืืžื•ื ื” ื›ื™ ืœื‘ืงื ื™ื ื ื—ื•ืชื™ื ืžืื ืฉื™ื ืขื ืจื™ื›ื•ื– ื’ื‘ื•ื” ื™ื•ืชืจ ืฉืœ ืžืœื ื™ืŸ ื‘ืขื•ืจื. ื›ืชื•ืฆืื” ืžื›ืš ื ืจื“ืคื•, ื ื”ืจื’ื• ื•ื‘ื•ืชืจื• ืœื‘ืงื ื™ื, ื•ืงื‘ืจื™ื ืฉืœ ืœื‘ืงื ื™ื ื ื—ืคืจื• ื•ื—ื•ืœืœื•. ืœื‘ืงื ื™ื ื”ื•ืฉืžื“ื• ื‘ื˜ืขื ื” ืฉื”ื ืžื‘ื™ืื™ื ืžื–ืœ ืจืข ื‘ืื–ื•ืจื™ื ืžืกื•ื™ืžื™ื. ื‘ืืคืจื™ืงื” ื”ืชื•ืคืขื” ืจื•ื•ื—ืช ื™ื•ืชืจ ืžืืฉืจ ื‘ืื–ื•ืจื™ื ืื—ืจื™ื ื‘ืขื•ืœื. ื‘ื”ืื™ื˜ื™ ื™ืฉ ื’ื ืžืกื•ืจืช ืฉืœ ื”ืชื™ื™ื—ืกื•ืช ื›ื–ื• ืœืœื‘ืงื ื™ื, ื‘ืฉื›ื™ื—ื•ืช ื’ื‘ื•ื”ื” ื‘ื”ืฉืคืขืชื• ืฉืœ ืคืจื ืกื•ืื” ื“ื•ื‘ืœื™ื”. ืื ืฉื™ื ืขืœ ื”ืกืคืงื˜ืจื•ื ื”ืื•ื˜ื™ืกื˜ื™ ื”ื™ื• ื‘ื“ืจืš ื›ืœืœ ืงื•ืจื‘ื ื•ืช ืœืจื“ื™ืคื”, ืœืื•ืจืš ื”ื”ื™ืกื˜ื•ืจื™ื” ื•ื’ื ื‘ืขื™ื“ืŸ ื”ื ื•ื›ื—ื™. ื‘ืงืžืจื•ืŸ ืžืืฉื™ืžื™ื ื™ืœื“ื™ื ืขื ืื•ื˜ื™ื–ื ื‘ื›ื™ืฉื•ืฃ ื•ืขื•ื ืฉื™ื ืื•ืชื ื‘ืขื™ื ื•ื™ื™ื ื•ืืฃ ื‘ืžื•ื•ืช. ื‘ื ื•ืกืฃ, ืžืฉืขืจื™ื ืฉืจื‘ื™ื ืžื”ื™ืœื“ื™ื ื”ื ื›ื™ื ืฉื ืจืฆื—ื• ื‘ืžื”ืœืš ืžื‘ืฆืข T4 ื‘ื’ืจืžื ื™ื” ื”ื ืืฆื™ืช ื”ื™ื• ืื•ื˜ื™ืกื˜ื™ื, ื•ื”ืคื›ื• ืื ืฉื™ื ืื•ื˜ื™ืกื˜ื™ื ืœืงื•ืจื‘ื ื•ืช ื”ืจืืฉื•ื ื™ื ืฉืœ ื”ืฉื•ืื”. ืœื”ื˜"ื‘ ื”ื•ืžื•ืกืงืกื•ืืœื™ื ื ืจื“ืคื• ื‘ื’ืจืžื ื™ื” ื”ื ืืฆื™ืช. ื‘ืชื—ื™ืœืช ื”ืžืื” ื”-21 ื ืจื“ืคื• ื•ืขื“ื™ื™ืŸ ื ืจื“ืคื™ื ื”ื•ืžื•ืกืงืกื•ืืœื™ื ื‘ืฆ'ืฆ'ื ื™ื” ื•ื‘ืžื“ื™ื ื” ื”ืืกืœืืžื™ืช. ืžืกืคืจ ืžื“ื™ื ื•ืช, ื‘ืžื™ื•ื—ื“ ืžื“ื™ื ื•ืช ื‘ืขื•ืœื ื”ืžืขืจื‘ื™, ื ืงื˜ื• ืืžืฆืขื™ื ื ื’ื“ ืืคืœื™ื” ืฉืœ ืžื™ืขื•ื˜ื™ื ืžื™ื ื™ื™ื, ื›ื•ืœืœ ื—ื•ืงื™ื ื ื’ื“ ืคืฉืขื™ ืฉื ืื” ื›ืœืคื™ ื”ื•ืžื•ืกืงืกื•ืืœื™ื ื•ืืคืœื™ื” ื‘ืžืงื•ื ื”ืขื‘ื•ื“ื”. ื—ืœืงื ืื™ืฉืจื• ื ื™ืฉื•ืื™ืŸ ื—ื“-ืžื™ื ื™ื™ื ืื• ืื™ื—ื•ื“ื™ื ืื–ืจื—ื™ื™ื ื‘ืžื˜ืจื” ืœื”ืขื ื™ืง ืœื–ื•ื’ื•ืช ื—ื“ ืžื™ื ื™ื™ื ืืช ืื•ืชืŸ ื”ื’ื ื•ืช ื•ื”ื˜ื‘ื•ืช ื›ืžื• ืœื–ื•ื’ื•ืช ื“ื• ืžื™ื ื™ื™ื. ื‘ืฉื ืช 2011 ืงื™ื‘ืœื• ื”ืื•ืžื•ืช ื”ืžืื•ื—ื“ื•ืช ืืช ื”ื—ืœื˜ืชืŸ ื”ืจืืฉื•ื ื” ืœื”ื›ืจื” ื‘ื–ื›ื•ื™ื•ืช ืœื”ื˜"ื‘, ื•ื‘ืฉื ืช 2015 ื”ืคื›ื• ื ื™ืฉื•ืื™ื ื—ื“ ืžื™ื ื™ื™ื ืœื—ื•ืงื™ื™ื ื‘ื›ืœ ืžื“ื™ื ื•ืช ืืจืฆื•ืช ื”ื‘ืจื™ืช. ืจืื• ื’ื ืงื™ืฉื•ืจื™ื ื—ื™ืฆื•ื ื™ื™ื ื”ืขืจื•ืช ืฉื•ืœื™ื™ื
========================================
[SOURCE: https://huggingface.co/models?pipeline_tag=text-to-video] | [TOKENS: 1084]
Models tencent/HunyuanVideo-1.5 Text-to-Video โ€ข Updated Dec 25, 2025 โ€ข 1.35k โ€ข โ€ข 575 Wan-AI/Wan2.2-TI2V-5B Text-to-Video โ€ข Updated Aug 7, 2025 โ€ข 4.81k โ€ข โ€ข 516 tencent/HunyuanVideo Text-to-Video โ€ข Updated Mar 6, 2025 โ€ข 1.15k โ€ข โ€ข 2.12k genmo/mochi-1-preview Text-to-Video โ€ข Updated Sep 4, 2025 โ€ข 4.93k โ€ข โ€ข 1.31k meituan-longcat/LongCat-Video Text-to-Video โ€ข Updated Oct 29, 2025 โ€ข 1.09k โ€ข โ€ข 439 CodeGoat24/Wan2.2-T2V-A14B-UnifiedReward-Flex-lora Text-to-Video โ€ข Updated 11 days ago โ€ข 164 โ€ข 8 Wan-AI/Wan2.1-T2V-14B Text-to-Video โ€ข Updated Mar 12, 2025 โ€ข 36.6k โ€ข โ€ข 1.47k Wan-AI/Wan2.2-T2V-A14B Text-to-Video โ€ข Updated Aug 7, 2025 โ€ข 3.57k โ€ข โ€ข 423 zai-org/CogVideoX-5b Text-to-Video โ€ข Updated Nov 23, 2024 โ€ข 33.5k โ€ข โ€ข 663 vrgamedevgirl84/Wan14BT2VFusioniX Text-to-Video โ€ข Updated Jun 21, 2025 โ€ข 603 QuantStack/Wan2.2-TI2V-5B-GGUF Text-to-Video โ€ข 5B โ€ข Updated Jul 31, 2025 โ€ข 11.6k โ€ข 137 QuantStack/Wan2.2-T2V-A14B-GGUF Text-to-Video โ€ข 14B โ€ข Updated Jul 29, 2025 โ€ข 125k โ€ข 222 lightx2v/Wan2.2-Lightning Text-to-Video โ€ข Updated Nov 13, 2025 โ€ข 69 โ€ข 594 GitMylo/Wan_2.2_nvfp4 Text-to-Video โ€ข Updated 16 days ago โ€ข 27 ali-vilab/text-to-video-ms-1.7b Text-to-Video โ€ข Updated Dec 1, 2023 โ€ข 11.4k โ€ข 651 calcuis/hunyuan-gguf Text-to-Video โ€ข 13B โ€ข Updated Dec 21, 2024 โ€ข 417 โ€ข 70 Wan-AI/Wan2.1-T2V-1.3B Text-to-Video โ€ข Updated Mar 1, 2025 โ€ข 12.8k โ€ข โ€ข 431 city96/Wan2.1-T2V-14B-gguf Text-to-Video โ€ข 14B โ€ข Updated Feb 26, 2025 โ€ข 20.5k โ€ข 184 calcuis/wan-1.3b-gguf Text-to-Video โ€ข 0.1B โ€ข Updated Aug 8, 2025 โ€ข 3.75k โ€ข 35 alibaba-pai/Wan2.1-Fun-1.3B-Control Text-to-Video โ€ข Updated Dec 11, 2025 โ€ข 1.42k โ€ข 110 Remade-AI/Vintage-VHS Text-to-Video โ€ข Updated Mar 27, 2025 โ€ข 17 โ€ข 3 Skywork/SkyReels-V2-DF-1.3B-540P Text-to-Video โ€ข 1B โ€ข Updated Apr 25, 2025 โ€ข 648 โ€ข 44 Skywork/SkyReels-V2-T2V-14B-720P Text-to-Video โ€ข Updated Apr 25, 2025 โ€ข 373 โ€ข 41 Lightricks/LTX-Video-0.9.7-dev Text-to-Video โ€ข Updated Jul 8, 2025 โ€ข 560 โ€ข โ€ข 20 Skywork/SkyReels-V2-DF-14B-540P-Diffusers Text-to-Video โ€ข Updated Aug 11, 2025 โ€ข 764 โ€ข 4 lightx2v/Wan2.1-T2V-14B-StepDistill-CfgDistill Text-to-Video โ€ข Updated Oct 17, 2025 โ€ข 130 APRIL-AIGC/UltraWan Text-to-Video โ€ข Updated Dec 11, 2025 โ€ข 8 โ€ข 27 Lightricks/LTX-Video-ICLoRA-pose-13b-0.9.7 Text-to-Video โ€ข Updated Jul 8, 2025 โ€ข 1.38k โ€ข 17 Wan-AI/Wan2.2-T2V-A14B-Diffusers Text-to-Video โ€ข Updated Aug 9, 2025 โ€ข 130k โ€ข โ€ข 110 alibaba-pai/Wan2.2-Fun-A14B-Control-Camera Text-to-Video โ€ข Updated Dec 11, 2025 โ€ข 3.44k โ€ข 34
========================================
[SOURCE: https://huggingface.co/docs/hub/spaces-zerogpu] | [TOKENS: 579]
Hub documentation Spaces ZeroGPU: Dynamic GPU Allocation for Spaces Hub and get access to the augmented documentation experience to get started Spaces ZeroGPU: Dynamic GPU Allocation for Spaces ZeroGPU is a shared infrastructure that optimizes GPU usage for AI models and demos on Hugging Face Spaces. It dynamically allocates and releases NVIDIA H200 GPUs as needed, offering: Unlike traditional single-GPU allocations, ZeroGPUโ€™s efficient system lowers barriers for developers, researchers, and organizations to deploy AI models by maximizing resource utilization and power efficiency. Using and hosting ZeroGPU Spaces Technical Specifications ZeroGPU supports two GPU sizes See GPU size selection to learn how to use sizes Compatibility ZeroGPU Spaces are designed to be compatible with most PyTorch-based GPU Spaces. While compatibility is enhanced for high-level Hugging Face libraries like transformers and diffusers, users should be aware that: Gradio: 4+ PyTorch: Almost all versions from 2.1.0 to latest are supported Python: Getting started with ZeroGPU To utilize ZeroGPU in your Space, follow these steps: This decoration process allows the Space to request a GPU when the function is called and release it upon completion. Note: The @spaces.GPU decorator is designed to be effect-free in non-ZeroGPU environments, ensuring compatibility across different setups. GPU size selection The default size used by @spaces.GPU is large (half H200). You can explicitly request a full H200 by specifying size="xlarge": Duration Management For functions expected to exceed the default 60-second of GPU runtime, you can specify a custom duration: This sets the maximum function runtime to 120 seconds. Specifying shorter durations for quicker functions will improve queue priority for Space visitors. @spaces.GPU also supports dynamic durations. Instead of directly passing a duration, simply pass a callable that takes the same inputs as your decorated function and returns a duration value: Compilation ZeroGPU does not support torch.compile, but you can use PyTorch ahead-of-time compilation (requires torch 2.8+) Check out this blogpost for a complete guide on ahead-of-time compilation on ZeroGPU. Usage Tiers GPU usage is subject to daily quotas, per account tier: Quota resets exactly 24 hours after your first GPU usage. Remaining quota directly impacts priority in ZeroGPU queues. Hosting Limitations By leveraging ZeroGPU, developers can create more efficient and scalable Spaces, maximizing GPU utilization while minimizing costs. Recommendations If your demo uses a large model, we recommend using optimizations like ahead-of-time compilation and flash-attention 3. You can learn how to leverage these with ZeroGPU in this post. These optimizations will help you to maximize the advantages of ZeroGPU hours and provide a better user experience. Feedback You can share your feedback on Spaces ZeroGPU directly on the HF Hub: https://huggingface.co/spaces/zero-gpu-explorers/README/discussions
========================================
[SOURCE: https://huggingface.co/spaces?hardware=zerogpu] | [TOKENS: 1080]
Spaces The AI App Directory Running on Zero MCP 830 Wan2.2 14B Preview ๐ŸŒ generate a video from an image with a text prompt r3gm 5 days ago generate a video from an image with a text prompt Running on Zero MCP 2.25k Z Image Turbo ๐Ÿ–ผ Generate high-quality images from text prompts in seconds mrfakename 19 days ago Generate high-quality images from text prompts in seconds Running on Zero Featured 108 SoulX-Singer ๐ŸŽค Generate singing voice from your lyrics Soul-AILab 11 days ago Generate singing voice from your lyrics Running on Zero Featured 1.48k Qwen3-TTS Demo ๐ŸŽ™ Generate custom speech from text, voice descriptions, or samples Qwen 5 days ago Generate custom speech from text, voice descriptions, or samples Running on Zero Featured 94 FireRed Image Edit 1.0 ๐ŸŒ FireRed-Image-Edit-1.0 FireRedTeam 5 days ago FireRed-Image-Edit-1.0 Running on Zero MCP Featured 873 Qwen-Image-Edit-2511-LoRAs-Fast ๐ŸŽƒ Demo of the Collection of Qwen Image Edit LoRAs prithivMLmods 3 days ago Demo of the Collection of Qwen Image Edit LoRAs Running on Zero MCP 381 Wan2.2 14B Fast Preview ๐ŸŒ generate a video from an image with a text prompt r3gm 15 days ago generate a video from an image with a text prompt Running on Zero Featured 422 ACE-Step v1.5 ๐ŸŽต Music Generation Foundation Model v1.5 ACE-Step 1 day ago Music Generation Foundation Model v1.5 Running on Zero Featured 1.07k TRELLIS.2 ๐Ÿข High-fidelity 3D Generation from images microsoft Dec 17, 2025 High-fidelity 3D Generation from images Running on Zero 42 Omni Video Factory ๐Ÿ† text to video, image to video, video extend FrameAI4687 5 days ago text to video, image to video, video extend Running on Zero MCP 232 LTX-2 Video [Turbo] ๐Ÿ”ฅ Fast high quality video with audio generation with FA3 alexnasa 13 days ago Fast high quality video with audio generation with FA3 Running on Zero Featured 1.45k Qwen Image Multiple Angles 3D Camera ๐ŸŽฅ Change the camera angle of a photo with AI multimodalart Jan 8 Change the camera angle of a photo with AI Running on Zero MCP Featured 528 FLUX.2 [Klein] 9B ๐Ÿ’ป Generate or edit images from text prompts with optional input images black-forest-labs Jan 16 Generate or edit images from text prompts with optional input images Running on Zero MCP 37 BitDance-14B-64x ๐Ÿš€ Open-source autoregressive model with binary visual tokens. shallowdream204 4 days ago Open-source autoregressive model with binary visual tokens. Running on Zero 322 NSFW Uncensored Adult Image ๐Ÿ“ˆ Based 'Z-IMAGE TURBO' Heartsync Jan 1 Based 'Z-IMAGE TURBO' Running on Zero Featured 80 Kugel Audio ๐Ÿ‘€ Generate natural-sounding speech in European languages with voice cloning multimodalart 16 days ago Generate natural-sounding speech in European languages with voice cloning Running on Zero MCP Featured 1.3k Dream-wan2-2-faster-Pro ๐ŸŽฅ generate a video from an image with a text prompt dream2589632147 Dec 24, 2025 generate a video from an image with a text prompt Running on Zero MCP Featured 2.79k Wan2.2 14B Fast ๐ŸŽฅ generate a video from an image with a text prompt zerogpu-aoti Dec 16, 2025 generate a video from an image with a text prompt Running on Zero MCP Featured 66 GLM OCR Demo ๐Ÿ“„ Multimodal OCR model for complex document understanding. prithivMLmods 3 days ago Multimodal OCR model for complex document understanding. Running on Zero MCP 185 Wan2.2 14B Fast ๐ŸŽฅ app_lora is base rahul7star 12 days ago app_lora is base Running on Zero MCP Featured 2k Qwen Image Edit Camera Control ๐ŸŽฌ Fast 4 step inference with Qwen Image Edit 2509 linoyts Jan 12 Fast 4 step inference with Qwen Image Edit 2509 Running on Zero 24 MOSS TTS ๐Ÿ† A simple gradio platform for demonstrating MOSS-TTS capabili OpenMOSS-Team 8 days ago A simple gradio platform for demonstrating MOSS-TTS capabili Running on Zero Featured 2.05k Hunyuan3D-2.1 ๐Ÿ‘ป Image-to-3D Generation tencent Aug 11, 2025 Image-to-3D Generation Running on Zero MCP Featured 222 Qwen Edit Any Pose ๐Ÿ•บ Edit any pose with Qwen Edit 2511 Any Pose LoRA linoyts Dec 29, 2025 Edit any pose with Qwen Edit 2511 Any Pose LoRA
========================================
[SOURCE: https://he.wikipedia.org/wiki/ื”ื—ื•ืง_ื”ื‘ื™ืŸ-ืœืื•ืžื™] | [TOKENS: 11394]
ืชื•ื›ืŸ ืขื ื™ื™ื ื™ื ืžืฉืคื˜ ื‘ื™ืŸ-ืœืื•ืžื™ ืคื•ืžื‘ื™ ืžืฉืคื˜ ื‘ื™ืŸ-ืœืื•ืžื™ ืคื•ืžื‘ื™ ื”ื•ื ืขื ืฃ ืžืฉืคื˜ื™ ื”ืขื•ืกืง ื‘ืคืขื•ืœื•ืชื™ื”ื ืฉืœ ื’ื•ืจืžื™ื ื‘ื–ื™ืจื” ื”ื‘ื™ืŸ-ืœืื•ืžื™ืช, ื›ื’ื•ืŸ ืžื“ื™ื ื•ืช, ืืจื’ื•ื ื™ื ื‘ื™ืŸ-ืœืื•ืžื™ื™ื ื•ืืฃ ืคืจื˜ื™ื ื”ื ื•ืฉืื™ื ื‘ืื—ืจื™ื•ืช ื‘ื™ืŸ-ืœืื•ืžื™ืช. ืชื—ื•ื ื–ื” ืžืกื“ื™ืจ ืืช ื”ื™ื—ืกื™ื ืฉื‘ื™ืŸ ืžื“ื™ื ื•ืช, ืืช ื”ืฉื™ืžื•ืฉ ื‘ื›ื•ื— ื•ืืช ื”ืžื ื’ื ื•ื ื™ื ื”ื‘ื™ืŸ-ืœืื•ืžื™ื™ื ืœื™ื™ืฉื•ื‘ ืกื›ืกื•ื›ื™ื. ื‘ืžืงืจื™ื ืžืกื•ื™ืžื™ื ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื”ืคื•ืžื‘ื™ ื’ื ืงื•ื‘ืข ื›ืœืœื™ื ื‘ื™ื—ืก ืœืื ืฉื™ื ืื• ืœื™ืฉื•ื™ื•ืช ืžืฉืคื˜ื™ื•ืช ืฉื•ื ื•ืช ื•ืžืืคืฉืจ ืœื™ื™ื—ืก ืœื”ื ืื—ืจื™ื•ืช ืคืœื™ืœื™ืช. ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื”ืคื•ืžื‘ื™ ื ืคืจื“ ืžืชื—ื•ื ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื”ืคืจื˜ื™. ื”ื‘ืกื™ืก ื”ื”ื™ืกื˜ื•ืจื™ ืœืžืฉืคื˜ ื‘ื™ืŸ-ืœืื•ืžื™ ืžืฉื—ืจ ื”ื”ื™ืกื˜ื•ืจื™ื” ื ื™ื”ืœื• ื™ืฉื•ื™ื•ืช ืžื“ื™ื ื™ื•ืช ื™ื—ืกื™ ื—ื•ืฅ ืขื ืžื“ื™ื ื•ืช ืื—ืจื•ืช, ื”ื—ืœื™ืคื• ืฉืœื™ื—ื™ื ื“ื™ืคืœื•ืžื˜ื™ื™ื ื•ื’ื™ื‘ืฉื• ื”ืกื›ืžื™ื ื”ื ื•ื’ืขื™ื ืœืฉื™ืชื•ืคื™ ืคืขื•ืœื” ื›ืœื›ืœื™ื™ื ื•ืฆื‘ืื™ื™ื, ื›ืจืชื• ื‘ืจื™ืชื•ืช ื•ื”ืกื›ื™ืžื• ืขืœ ืžื ื’ื ื•ื ื™ื ืœื™ื™ืฉื•ื‘ ืกื›ืกื•ื›ื™ื ื‘ื›ืชื‘ ื•ื‘ืขืœ ืคื”. ื‘ื”ื“ืจื’ื” ื”ืชืคืชื—ื• ื›ืœืœื™ื ืืฉืจ ื ืฉืื• ืื•ืคื™ ื˜ืงืกื™ ื•ืžื—ื™ื™ื‘ ื‘ื™ื—ืก ืœื›ื™ื‘ื•ื“ ื”ื”ืกื›ืžื™ื ื•ืœื”ืคืจืชื. ื™ื“ื•ืขื™ื ืžืงืจื™ื ืฉืœ ื“ื™ืคืœื•ืžื˜ื™ื” ื•ื”ืกื›ืžื™ื ื‘ื™ืŸ-ืœืื•ืžื™ื™ื ื›ื‘ืจ ื‘ืžื–ืจื— ื”ืงื“ื•ื, ื•ื‘ืžืงืจื™ื ืžืกื•ื™ืžื™ื ืืฃ ืฉืจื“ื• ืขื“ื•ื™ื•ืช ืœื”ืกื›ืžื™ื ืืœื•. ื’ื ื‘ืชื "ืš ืจื•ืื™ื ืขื“ื•ื™ื•ืช ืœืžืฉืคื˜ ื‘ื™ืŸ-ืœืื•ืžื™, ืœื“ื•ื’ืžื” ื‘ื•ื•ื™ื›ื•ื— ื”ืืจื•ืš ื‘ื™ืŸ ื™ืคืชื—, ืฉื•ืคื˜ ื™ืฉืจืืœื™ ื‘ืชืงื•ืคืช ื”ืฉื•ืคื˜ื™ื, ืœื‘ื™ืŸ ืžืœืš ื‘ื ื™ ืขืžื•ืŸ ืœื’ื‘ื™ ื”ืžืขืžื“ ื”ืžืฉืคื˜ื™ ืฉืœ ืฉื˜ื— ืžืกื•ื™ื ื‘ืขื‘ืจ ื”ื™ืจื“ืŸ ื”ืžื–ืจื—ื™ (ืกืคืจ ืฉื•ืคื˜ื™ื, ืคืจืง ื™"ื). ืžื”ื•ื•ื™ื›ื•ื— ืขื•ืœื” ื‘ื‘ื™ืจื•ืจ ื›ื™ ื”ื™ื• ืžืงื•ื‘ืœื™ื ืขืœ ืฉื ื™ ื”ืฆื“ื“ื™ื ืขืงืจื•ื ื•ืช ืžืกื•ื™ืžื™ื ืฉืœ ืฆื“ืง. ื”ื ื™ืกื™ื•ื ื•ืช ื”ืคื•ืจืžืœื™ื™ื ื”ืจืืฉื•ื ื™ื ืœืžืกื“ ืืช ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื”ื—ืœื• ื‘ืื™ืจื•ืคื” ื‘ืชืงื•ืคืช ื”ืจื ืกืื ืก. ื‘ื™ืžื™ ื”ื‘ื™ื ื™ื™ื ื”ื›ื ืกื™ื™ื” ื”ื™ื™ืชื” ื”ืžื’ืฉืจืช ื”ืขื™ืงืจื™ืช ื‘ืกื›ืกื•ื›ื™ื ื‘ื™ืŸ-ืœืื•ืžื™ื™ื. ื‘ื–ืžืŸ ื•ืขื™ื“ืช ืงื•ื ืกื˜ื ืฅ (1414), ืคืœื•ื•ืœ ื•ื•ืœื•ื“ืงื•ื•ื™ืง โ€“ ื”ืจืงื˜ื•ืจ ืฉืœ ื”ืื•ื ื™ื‘ืจืกื™ื˜ื” ื”ื™ื’ืœื•ื ื™ืช (ืงืจืงื•ื‘, ืคื•ืœื™ืŸ) โ€“ ืชืื•ืœื•ื’, ืขื•ืจืš ื“ื™ืŸ ื•ื“ื™ืคืœื•ืžื˜, ื”ืฆื™ื’ ืืช ื”ืชืื•ืจื™ื” ื›ื™ ืœื›ืœ ื”ืื•ืžื•ืช, ื’ื ื”ืคื’ืื ื™ื•ืช, ื™ืฉ ื–ื›ื•ืช ืœืžืžืฉืœ ืขืฆืžืื™ ื•ืœื—ื™ื™ื ื‘ืฉืœื•ื ื‘ืืจืฆื. ื‘ืžืื•ืช ื”-16 ื•ื”-17 ื‘ืื™ืจื•ืคื”, ื”ื›ื ืกื™ื™ื” ืื™ื‘ื“ื” ืืช ื”ืฉืคืขืชื” ื”ื™ืฉื™ืจื” ื‘ืขื ื™ื™ื ื™ื ื‘ื™ืŸ-ืœืื•ืžื™ื™ื, ื›ืืฉืจ ืžื“ื™ื ื•ืช ืงืชื•ืœื™ื•ืช ื•ืคืจื•ื˜ืกื˜ื ื˜ื™ื•ืช ื”ื•ืคื™ืขื• ื•ื ืื‘ืงื• ื‘ื™ื ื™ื”ืŸ ืขืœ ื”ืฉืœื˜ื•ืŸ. ื‘ืชื—ื™ืœืช ื”ืžืื” ื”-17, ืืคืฉืจ ื”ื™ื” ืœื•ืžืจ ื‘ืื•ืคืŸ ืžื›ืœื™ืœ ื›ืžื” ื“ื‘ืจื™ื ืœื’ื‘ื™ ื”ืžืฆื‘ ื”ืคื•ืœื™ื˜ื™: ื™ืฉ ื”ื˜ื•ืขื ื™ื ื›ื™ ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื”ืชืคืชื— ืขืœ ืžื ืช ืœื”ืชืžื•ื“ื“ ืขื ื”ืžื“ื™ื ื•ืช ื”ื—ื“ืฉื•ืช ืฉืงืžื•, ืื—ืจื™ื ื˜ื•ืขื ื™ื ื›ื™ ื”ื—ืœืœ ืฉื ื•ืฆืจ ืœืื—ืจ ื™ืจื™ื“ืช ื”ื›ื ืกื™ื™ื” ื•ื”ืืคื™ืคื™ื•ืจื•ืช ื”ื ืฉื”ื‘ื™ืื• ืœืฆื•ืจืš ื‘ื—ื•ืงื™ื ื›ืœืœื™ื™ื ื—ื“ืฉื™ื[ื“ืจื•ืฉ ืžืงื•ืจ]. ืคืจื ืกื™ืกืงื• ื“ื” ื•ื™ื˜ื•ืจื™ื” (Vitoria), ืคืจื•ืคืกื•ืจ ื“ื•ืžื™ื ื™ืงื ื™ ืœืชืื•ืœื•ื’ื™ื” ื‘ืื•ื ื™ื‘ืจืกื™ื˜ืช ืกืœืžื ืงื” ื”ืจืฆื” ืขืœ ื–ื›ื•ื™ื•ืชื™ื”ื ืฉืœ ื”ื™ืœื™ื“ื™ื. ื”ื•ื ืขืฉื” ื–ืืช ื‘ื–ืžืŸ ืฉืกืคืจื“ ื”ื™ื™ืชื” ื‘ืฉื™ื ืชืคืืจืชื”, ืœืื—ืจ ื”ื›ื™ื‘ื•ืฉ ื”ืืœื™ื ืฉืœ ืคืจื• ื‘ึพ1536. ืงืจืœ ื”ื—ืžื™ืฉื™, ืงื™ืกืจ ื”ืื™ืžืคืจื™ื” ื”ืจื•ืžื™ืช ื”ืงื“ื•ืฉื”, ื”ืชื ื’ื“ ืœืคืจื•ืคืกื•ืจ, ืืš ื‘ึพ1542 ื—ื•ืงื™ื ื—ื“ืฉื™ื ืงื‘ืขื• ื›ื™ ื”ื™ืœื™ื“ื™ื ืžื•ื’ื ื™ื ืขืœ ื™ื“ื™ ื”ื›ืชืจ ื”ืกืคืจื“ื™. ื•ื™ื˜ื•ืจื™ื” ื‘ื“ืจืš ื›ืœืœ ื ื—ืฉื‘ ืœืžื™ื™ืกื“ ืฉืœ ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื”ืžื•ื“ืจื ื™. ื”ื ื–ื™ืจ ื”ืฆืจืคืชื™ ืืžืจื™ืง ืงืจื•ืก (Crucรฉ ,1590โ€“1648) ื—ืฉื‘ ื›ื™ ื ืฆื™ื’ื™ื ืฉืœ ื›ืœ ื”ืžื“ื™ื ื•ืช ืฉื™ื™ืคื’ืฉื• ื‘ืžืงื•ื ืื—ื“ ืขืœ ืžื ืช ืœื“ื•ืŸ ื‘ืกื›ืกื•ื›ื™ื”ืŸ, ื™ื•ื›ืœื• ืœืžื ื•ืข ืžืœื—ืžื” ื•ืœื”ื‘ื™ื ืœืฉืœื•ื. ื‘-1623, ื”ื•ื ื”ืฆื™ืข ืืช ื”ืจืขื™ื•ืŸ ื‘"ื”ืกื™ื ืื™ื ื™ื ื”ื—ื“ืฉื™ื" (Nouveau Cynรฉe), ื›ืฉื”ื•ื ืžืฆื™ืข ืืช ื•ื ืฆื™ื” ื›ืขื™ืจ ืฉื‘ื” ื™ืคื’ืฉื• ื”ื ืฆื™ื’ื™ื, ื•ื”ืฆื™ืข ื›ื™ ื”ืืคื™ืคื™ื•ืจ ื”ื•ื ืฉื™ืฉื‘ ื‘ืจืืฉ ื”ืืกืคื”. ื‘ื–ืžืŸ ืžืœื—ืžืช ืฉืœื•ืฉื™ื ื”ืฉื ื™ื (1618โ€“1648), ืจืขื™ื•ืŸ ื›ื–ื” ืœื ื”ื™ื” ืžืงื•ื‘ืœ ืขืœ ื”ืžื“ื™ื ื•ืช ื”ืคืจื•ื˜ืกื˜ื ื˜ื™ื•ืช. ื”ื•ื ื’ื ืืžืจ ื›ื™ ื™ืฉ ืœื‘ื˜ืœ ืืช ื”ืฆื‘ืื•ืช ื•ืœื™ื™ืกื“ ื‘ื™ืช ื“ื™ืŸ ื‘ื™ืŸ-ืœืื•ืžื™. ืืฃ ืฉื”ืจืขื™ื•ื ื•ืช ืฉืœื• ืœื‘ื™ื˜ื•ืœ ื”ืฆื‘ืื•ืช ืœื ื ืœืงื—ื• ื‘ืจืฆื™ื ื•ืช, ืœืงืจื•ืก ื”ื™ื” ื‘ืขืœ ื”ื”ื‘ื ื” ื”ืžื•ืงื“ืžืช ื•ื”ืจืืฉื•ืŸ ืฉื”ืฆื™ืข ื›ื™ ืืจื’ื•ื ื™ื ื‘ื™ืŸ-ืœืื•ืžื™ื™ื ื”ื›ืจื—ื™ื™ื ืขืœ ืžื ืช ืœืคืชื•ืจ ืกื›ืกื•ื›ื™ื ื‘ื™ืŸ-ืœืื•ืžื™ื™ื. ื”ื•ื’ื• ื’ืจื•ื˜ื™ื•ืก (1583-1645) ื”ื™ื” ื”ื•ืžื ื™ืกื˜ ื•ืžืฉืคื˜ืŸ ื”ื•ืœื ื“ื™, ืฉื ื—ืฉื‘ ื›ืžืจื›ื–ื™ ื‘ื”ืชืคืชื—ื•ืช ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™. ื”ื•ื ื ืขืฉื” ืœืขื•ืจืš ื“ื™ืŸ ื›ืฉื”ื™ื” ื‘ืŸ ื—ืžืฉ ืขืฉืจื”. ื”ื•ื ื ืฉืคื˜ ืœืžืืกืจ ืขื•ืœื ืœืื—ืจ ืฉื”ืชื ื’ื“ ืœืžืื•ืจื™ืฅ, ื ืกื™ืš ืื•ืจื ื–', ื‘ืžืฉืคื˜, ืืš ื”ื•ื ื‘ืจื— ืœืคืจื™ื–. ื‘ืฆืจืคืช ืคื™ืชื— ืืช ืจืขื™ื•ื ื•ืชื™ื• ืขืœ ืžืฉืคื˜ ื‘ื™ืŸ-ืœืื•ืžื™ ื‘ืกืคืจื• "Mare Liberum" ("ื™ื ื—ื•ืคืฉื™"), ืฉื‘ื• ืชืงืฃ ืืช ื”ื ื™ืกื™ื•ื ื•ืช ืฉืœ ืื ื’ืœื™ื”, ืกืคืจื“ ื•ืคื•ืจื˜ื•ื’ืœ ืœืฉืœื•ื˜ ื‘ื—ืœืงื™ื ืžื”ื™ื. ื”ื•ื ื”ืชืคืจืกื ื‘ืขื•ืœื ื‘-1625 ืขื ืคืจืกื•ื ืกืคืจื• "ืžืฉืคื˜ ื”ืžืœื—ืžื” ื•ื”ืฉืœื•ื" (De Jure Belli ac Pacis), ื•ืกืคืจ ื–ื” ื ืขืฉื” ืœื˜ืงืกื˜ ื”ืงื•ื‘ืข ืœืžืฉืคื˜ ื‘ื™ืŸ-ืœืื•ืžื™. ื”ื•ื ืคื•ืจืกื ืจืง ืฉื ืชื™ื™ื ืœืื—ืจ "The New Cyneas". ืจื‘ื™ื ืžื”ืชื›ื ื™ื ืฉืœ ื’ืจื•ื˜ื™ื•ืก ื”ื’ื™ืขื• ืžื”ืชื "ืš ื•ืžื”ื”ื™ืกื˜ื•ืจื™ื” ื”ืงืœืืกื™ืช (ืžืžืฉ ื›ืžื• ืชืื•ืจื™ื™ืช ื”ืžืœื—ืžื” ืฉืœ ืื•ื’ื•ืกื˜ื™ื ื•ืก). ื‘ื—ื™ื‘ื•ืจื• ื”ื•ื ืœื ื”ืชื ื’ื“ ืœืžืœื—ืžื” ื›ื›ืœื™ ืคื•ืœื™ื˜ื™, ืืœื ืจืง ื”ื’ื‘ื™ืœ ืืช ื”ืžืœื—ืžื” ืœืžืงืจื™ื ืžืกื•ื™ืžื™ื ืฉื‘ื”ื ื”ื™ื ืœื’ื™ื˜ื™ืžื™ืช. ื”ื•ื ืคื™ืชื— ืืช ืชืื•ืจื™ื™ืช ื”ืžืœื—ืžื” ื”ืฆื•ื“ืงืช, ื•ื”ื’ื“ื™ืจ ืžืกืคืจ ืงืจื™ื˜ืจื™ื•ื ื™ื ืฉืขืฉื•ื™ื™ื ืœื”ืฆื™ื“ืง ืžืœื—ืžื”: ืžื“ื™ื ืื™ื ื‘ื ื™ ื”ื–ืžืŸ ื”ืืžื™ื ื• ื›ื™ ืœื ื ื™ืชืŸ ืœื”ืคืกื™ืง ืืช ื”ืžืœื—ืžื”, ื•ืœื›ืŸ ื”ื ื”ืชื›ื•ื ื ื• ืœื”. ื•ืขื™ื“ืช ื–'ื ื‘ื” ื”ืจืืฉื•ื ื” ื‘-1864 ื›ื™ื ืกื” ืืช ื”ืžื“ื™ื ื•ืช ื”ืื™ืจื•ืคืื™ื•ืช ืขืœ ืžื ืช ืœืžืกื“ ืœืจืืฉื•ื ื” ืืช ื—ื•ืงื™ ื”ืžืœื—ืžื” ื‘ืื™ืจื•ืคื” ื‘ืืžื ื” ื‘ื™ืŸ-ืœืื•ืžื™ืช. ื”ืืžื ื” ื”ืจืืฉื•ื ื” ืขืกืงื” ื‘ื˜ื™ืคื•ืœ ื‘ืคืฆื•ืขื™ ืžืœื—ืžื” ื•ื™ื™ืกื“ื” ืืช ืืจื’ื•ืŸ ื”ืฆืœื‘ ื”ืื“ื•ื ื”ื‘ื™ืŸ-ืœืื•ืžื™. ื‘ื”ืžืฉืš ื ื—ืชืžื• ืฉืœื•ืฉ ืืžื ื•ืช ื ื•ืกืคื•ืช, ื”ืืžื ื” ื”ืฉื ื™ื™ื” ื‘-1906 ื”ื—ื™ืœื” ืืช ืขืงืจื•ื ื•ืช ื”ืืžื ื” ื”ืจืืฉื•ื ื” ื‘ืœื•ื—ืžื” ื™ืžื™ืช, ื”ืฉืœื™ืฉื™ืช ื‘-1929 ืขืกืงื” ื‘ื˜ื™ืคื•ืœ ื‘ืฉื‘ื•ื™ื™ ืžืœื—ืžื”, ื•ื”ืจื‘ื™ืขื™ืช ื‘-1949 ืขืกืงื” ื‘ื˜ื™ืคื•ืœ ื‘ืื•ื›ืœื•ืกื™ื™ื” ืื–ืจื—ื™ืช ื‘ื–ืžืŸ ืžืœื—ืžื”. ื”ืืžื ื•ืช ื ื—ืชืžื• ืขืœ ื™ื“ื™ ื›ืœ ืžื“ื™ื ื•ืช ื”ืขื•ืœื, ื•ื”ืŸ ื ื—ืฉื‘ื•ืช ื›ื‘ืขืœื•ืช ืชื•ืงืฃ ืžืฉืคื˜ื™ ืžื—ื™ื™ื‘ ื•ืžืฉืงืคื•ืช ืžืฉืคื˜ ืžื ื”ื’ื™. ื›ื—ืœืง ืžืœืงื—ื™ ืžืœื—ืžืช ื”ืขื•ืœื ื”ืจืืฉื•ื ื”, ืžื“ื™ื ื•ืช ื”ืขื•ืœื ื”ื—ืœื™ื˜ื• ืœื”ืงื™ื ื’ื•ืฃ ื‘ื™ืŸ-ืœืื•ืžื™ ืืฉืจ ืžื˜ืจืชื• ืฉืžื™ืจื” ืขืœ ื”ืฉืœื•ื ื”ื‘ื™ืŸ-ืœืื•ืžื™ ืฉื™ื”ื•ื•ื” ืžื ื’ื ื•ืŸ ืžืกื•ื“ืจ ืœื™ื™ืฉื•ื‘ ืกื›ืกื•ื›ื™ื. ื•ืขื™ื“ืช ื”ืฉืœื•ื ื‘ืคืจื™ื– (1919), ื™ื™ืฆื’ื” ืืช ื”ื ื™ืกื™ื•ืŸ ื”ืจื—ื‘ ื”ืจืืฉื•ืŸ ืœื™ืฆื•ืจ ื”ืกื›ืžื™ื ื’ืœื•ื‘ืœื™ื™ื ื•ืจื‘-ืฆื“ื“ื™ื™ื. ื ืฉื™ื ืืจืฆื•ืช ื”ื‘ืจื™ืช, ื•ื•ื“ืจื• ื•ื™ืœืกื•ืŸ, ื”ืฆื™ืข ืœื”ืงื™ื ืืช ืืจื’ื•ืŸ ื—ื‘ืจ ื”ืœืื•ืžื™ื, ืืฉืจ ื”ื•ืงื ื‘-1920. ืืจื’ื•ืŸ ื–ื” ื›ืฉืœ ื‘ืžื˜ืจืชื• ืœืžื ื•ืข ืžืœื—ืžืช ืขื•ืœื ื ื•ืกืคืช, ื‘ื™ืŸ ื”ื™ืชืจ ืขืงื‘ ื—ื•ืกืจ ืฉื™ืชื•ืฃ ื”ืคืขื•ืœื” ืฉืœ ืžื“ื™ื ื•ืช ืžื•ื‘ื™ืœื•ืช ื‘ืขื•ืœื, ื‘ื™ื ื™ื”ืŸ ืืจืฆื•ืช ื”ื‘ืจื™ืช. ื›ืืฉืจ ืžืœื—ืžืช ื”ืขื•ืœื ื”ืฉื ื™ื™ื” ืคืจืฆื”, ื—ื‘ืจ ื”ืœืื•ืžื™ื ื”ืชื’ืœื” ื›ื’ื•ืฃ ื—ืกืจ ืชื•ืขืœืช ืœื™ื™ืฉื•ื‘ ืกื›ืกื•ื›ื™ื ื•ื”ื‘ื˜ื—ืช ื”ืฉืœื•ื ื”ื‘ื™ืŸ-ืœืื•ืžื™, ื•ืขืœื” ื”ืฆื•ืจืš ื‘ื”ืงืžืช ื’ื•ืฃ ื—ื“ืฉ ืืฉืจ ื™ื•ื›ืœ ืœื”ืชืžื•ื“ื“ ื‘ืขืชื™ื“ ื‘ื”ืฆืœื—ื” ืขื ืืชื’ืจื™ ื”ืžืขืจื›ืช ื”ื‘ื™ืŸ-ืœืื•ืžื™ืช. ื‘ึพ1 ื‘ื™ื ื•ืืจ 1942, ื ืฉื™ื ืืจืฆื•ืช ื”ื‘ืจื™ืช ืคืจื ืงืœื™ืŸ ื“ืœืื ื• ืจื•ื–ื•ื•ืœื˜ ืคืจืกื ืืช "ื”ืฆื”ืจืช ื”ืื•ืžื•ืช ื”ืžืื•ื—ื“ื•ืช" ื‘ืฉืžืŸ ืฉืœ 26 ืžื“ื™ื ื•ืช ืฉื”ืฆื”ื™ืจื• ื›ื™ ื”ืŸ ื™ืœื—ืžื• ื›ื ื’ื“ ืžื“ื™ื ื•ืช ื”ืฆื™ืจ. ืขื•ื“ ืœืคื ื™ ืกื•ืฃ ื”ืžืœื—ืžื”, ื ืฆื™ื’ื™ื ืฉืœ 50 ืžื“ื™ื ื•ืช ื ืคื’ืฉื• ื‘ืกืŸ ืคืจื ืกื™ืกืงื• ืขืœ ืžื ืช ืœื ืกื— ืืช ื”ืฆ'ืจื˜ืจ ืฉืœ ื’ื•ืฃ ื‘ื™ืŸ-ืœืื•ืžื™ ืฉื™ื—ืœื™ืฃ ืืช ื—ื‘ืจ ื”ืœืื•ืžื™ื, ื•ื‘ึพ24 ื‘ืื•ืงื˜ื•ื‘ืจ 1945 ื”ื•ืงื ืืจื’ื•ืŸ ื”ืื•ืžื•ืช ื”ืžืื•ื—ื“ื•ืช. ื”ืื•"ื ื•ื’ื•ืคื™ื• ืžื”ื•ื•ื™ื ืžืื– ืืช ื”ืžื•ืงื“ ื”ื—ืฉื•ื‘ ื‘ื™ื•ืชืจ ืœืžืขืจื›ื•ืช ื™ื—ืกื™ื ื“ื™ืคืœื•ืžื˜ื™ื•ืช, ืœื™ื™ืฉื•ื‘ ืกื›ืกื•ื›ื™ื ื•ืœืงื‘ื™ืขืช ืชื•ื›ื ื• ืฉืœ ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™, ื•ืžื’ื™ืœืช ื”ืื•ืžื•ืช ื”ืžืื•ื—ื“ื•ืช ืžื”ื•ื•ื” ืืช ื”ืฆ'ืจื˜ืจ ืฉืœื”ื. ื”ื”ื™ืงืฃ ืฉืœ ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื”ืขืจืš ื•ื”ืกืžื›ื•ืช ืฉืœ ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ืชืœื•ื™ ืœื—ืœื•ื˜ื™ืŸ ื‘ื”ืฉืชืชืคื•ืช ืžืจืฆื•ืŸ ืฉืœ ืžื“ื™ื ื•ืช ื‘ื™ืฆื™ืจื”, ืฉืžื™ืจื” ื•ืื›ื™ืคื” ืฉืœื•. ืืฃ ืขืœ ืคื™ ืฉื™ืฉ ืžืงืจื™ื ื™ื•ืฆืื™ ื“ื•ืคืŸ, ืจื•ื‘ ื”ืžื“ื™ื ื•ืช ืžืงื‘ืœื•ืช ืขืœื™ื”ืŸ ื”ืชื—ื™ื™ื‘ื•ื™ื•ืช ืžืฉืคื˜ื™ื•ืช ืœืžื“ื™ื ื•ืช ืื—ืจื•ืช ืžืชื•ืš ืื™ื ื˜ืจืก ืขืฆืžื™ ื•ืœื ืžืชื•ืš ืฆื™ื•ืช ืœื—ื•ืงื™ื ื’ื‘ื•ื”ื™ื ื™ื•ืชืจ ืžืฉืœ ื”ืžื“ื™ื ื” ืขืฆืžื”. ื”ื™ื™ืกื•ื“ ืฉืœ ื”ืื•ืžื•ืช ื”ืžืื•ื—ื“ื•ืช ื ืชืŸ ืืคืฉืจื•ืช ืœืงื”ื™ืœื” ื”ื‘ื™ืŸ-ืœืื•ืžื™ืช ืœืื›ื•ืฃ ืืช ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ืขืœ ืžื“ื™ื ื•ืช ื”ื—ื‘ืจื•ืช ื‘ืื•"ื ืฉืžืคืจื•ืช ืืช ื”ืฆ'ืจื˜ืจ ืฉืœื•. ื‘ืขื‘ืจ, ื ื—ืฉื‘ื• ืžื“ื™ื ื•ืช ืœืฆื“ื“ื™ื ื”ื™ื—ื™ื“ื™ื ืœืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™, ืœืคื™ ืชืื•ืจื™ื™ืช "ื›ื“ื•ืจื™ ื”ื‘ื™ืœื™ืืจื“". ื‘ืžืื” ื”ืฉื ื™ื ื”ืื—ืจื•ื ื•ืช ื’ื“ืœ ืžืื•ื“ ืžืกืคืจ ื”ืืจื’ื•ื ื™ื ื”ื‘ื™ืŸ-ืœืื•ืžื™ื™ื ื•ื”ื™ืฉื•ื™ื•ืช ื”ืœื-ืžื“ื™ื ืชื™ื•ืช ื‘ื–ื™ืจื” ื”ื‘ื™ืŸ-ืœืื•ืžื™ืช ื•ื›ื™ื•ื, ื’ื ื”ื ืžื•ื›ืจื™ื ื›ืฆื“ื“ื™ื ืจืœื•ื•ื ื˜ื™ื™ื. ืคื™ืจื•ืฉื™ื ืžื”ืขืช ื”ืื—ืจื•ื ื” ืฉืœ ืžืฉืคื˜ ื–ื›ื•ื™ื•ืช ื”ืื“ื ื”ื‘ื™ืŸ-ืœืื•ืžื™, ืžืฉืคื˜ ื”ื•ืžื ื™ื˜ืจื™ ื‘ื™ืŸ-ืœืื•ืžื™ ื•ืžืฉืคื˜ ืžืกื—ืจื™ ื‘ื™ืŸ-ืœืื•ืžื™ ื›ืœืœื• ืชืื’ื™ื“ื™ื ื•ื—ื‘ืจื•ืช, ื•ืืฃ ื™ื—ื™ื“ื™ื. ืžื—ืœื•ืงื•ืช ืขืงืจื•ื ื™ื•ืช ื‘ืชื•ืš ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื›ืขื ื™ื™ืŸ ืคื™ืœื•ืกื•ืคื™, ืคื•ืœื™ื˜ื™ ื•ื—ื•ืงืชื™, ืžื“ื™ื ื•ืช ืฉื•ืื‘ื•ืช ืืช ื”ืื•ื˜ื•ื ื•ืžื™ื” ืฉืœื”ืŸ ืžืœื’ื™ื˜ื™ืžืฆื™ื” ืคื ื™ืžื™ืช ื•ืœื ื‘ืฉืœ ื”ืกื›ืžื” ืฉืœ ื”ืงื”ื™ืœื” ื”ื‘ื™ืŸ-ืœืื•ืžื™ืช. ืžื“ื™ื ื•ืช ื™ื›ื•ืœื•ืช, ืœืคื™ื›ืš, ืœื”ืกื›ื™ื ืœื”ืชื—ื™ื™ื‘ ื‘ืื•ืคืŸ ื”ืชื ื“ื‘ื•ืชื™ ืœืžืฉืคื˜ ื‘ื™ืŸ-ืœืื•ืžื™, ืืš ื”ืŸ ืœื ืชืงื‘ืœื ื” ืชื”ืœื™ืš ืžืฉืคื˜ื™ ืžื—ื•ืฅ ืœื”ืกื›ืžืชืŸ ืฉืœื”ืŸ. ืœืคื™ื›ืš, ื”ืื™ื ื˜ืจืกื™ื ืฉืœื”ืŸ ื”ื ืฉื™ืงื‘ืขื• ื›ื™ืฆื“ ื”ื ืžืคืจืฉื™ื ืืช ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™. ืžืœื•ืžื“ื™ื ื•ืžื ื”ื™ื’ื™ื ืคื•ืœื™ื˜ื™ื™ื ืžืกื•ื™ืžื™ื ื˜ืขื ื• ื‘ื–ืžืŸ ื”ืื—ืจื•ืŸ[ื“ืจื•ืฉ ืžืงื•ืจ], ื›ื™ ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื›ื‘ืจ ื”ืชืคืชื— ืœืžืฆื‘ ืฉื‘ื• ื”ื•ื ืขืฆืžืื™ ืžื”ื”ืกื›ืžื” ืฉืœ ื”ืžื“ื™ื ื•ืช. ื™ืฉ ื ื˜ื™ื™ื” ื’ื•ื‘ืจืช ืœืฉืคื™ื˜ืช ื”ืžืขืฉื™ื ื”ืคื ื™ืžื™ื™ื ืฉืœ ืžื“ื™ื ื•ืช ืœืื•ืจ ื”ืกื˜ื ื“ืจื˜ื™ื ืฉืœ ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™. ื™ืฉื ืŸ ืžื“ื™ื ื•ืช, ื•ื‘ื™ื™ื—ื•ื“ ืืจืฆื•ืช ื”ื‘ืจื™ืช, ื”ืžืชื ื’ื“ื•ืช ื‘ืื•ืคืŸ ื ื—ืจืฅ ืœืคื™ืจื•ืฉ ื”ื–ื”, ื•ื˜ื•ืขื ื•ืช ื›ื™ ื”ืขืจืš ื”ื“ื•ืžื™ื ื ื˜ื™ ื”ื•ื ืจื™ื‘ื•ื ื•ืช. ื‘ื ื•ืกืฃ, ื™ืฉื ื ืื ืฉื™ ืืงื“ืžื™ื” ืฉืจื•ืื™ื ื”ื™ื•ื ืืช ืชื”ืœื™ื›ื™ ื”ืฉืคื™ื˜ื” ื•ื”ื—ืงื™ืงื” ื‘ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื›ืžืงื‘ื™ืœื™ื ืœืืœื” ืฉืœ ื”ืžืฉืคื˜ ื”ืื–ืจื—ื™ ื”ืคื ื™ืžื™. ืฉื•ื‘, ื”ืžืชื ื’ื“ื™ื ื˜ื•ืขื ื™ื ื›ื™ ืจืง ื”ืกื›ืžื” ืฉืœ ื”ืžื“ื™ื ื•ืช ื™ื›ื•ืœื” ืœื›ืคื•ืช ืขืœื™ื”ืŸ ืืช ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™. ืžืงื•ืจื•ืช ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ืกืขื™ืฃ 38 ืœื—ื•ืงืช ื‘ื™ืช ื”ื“ื™ืŸ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ืœืฆื“ืง (ICJ) ืงื•ื‘ืข ืืช ืžืงื•ืจื•ืช ื”ืžืฉืคื˜ ืขืœื™ื”ื ื™ืกืชืžืš ื‘ื‘ื•ืื• ืœื™ื™ืฉื‘ ืกื›ืกื•ื›ื™ื ื‘ื™ืŸ-ืœืื•ืžื™ื™ื. ื—ืžืฉืช ืžืงื•ืจื•ืช ืืœื”, ื”ืžื“ื•ืจื’ื™ื ืœืคื™ ืกื“ืจ ื—ืฉื™ื‘ื•ืชื, ืžื”ื•ื•ื™ื ืืช ืจืฉื™ืžืช ื”ืžืงื•ืจื•ืช ื”ืžืงื•ื‘ืœืช ืœืงื‘ื™ืขืช ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™. ื‘ืืฉืจ ืœืžืงื•ืจื•ืช, ืงื™ื™ืžื™ื ืžืกืคืจ ื“ื’ืฉื™ื. ืจืืฉื™ืช, ืœืคื™ ืก' 38 ืœื—ื•ืงืช ื‘ื™ืช ื”ื“ื™ืŸ, ืฉืœื•ืฉืช ื”ืžืงื•ืจื•ืช ื”ืจืืฉื•ื ื™ื ืžื›ื•ื ื™ื "ืžืงื•ืจื•ืช ืจืืฉื•ื ื™ื™ื" - ื”ืžื—ื™ื™ื‘ื™ื ื‘ืžื™ืฉื•ืจ ื”ื‘ื™ืŸ-ืœืื•ืžื™. ืืžื ื ื™ืฉ ืœืจืื•ืช ื‘ื”ื ื›ื”ื™ืจืจื›ื™ื™ื, ืืš ืœื”ื ืžืฉืžืขื•ืช ื–ื”ื”. ืฉื ื™ืช, ื‘ื”ืชืื ืœืก' 59 ืœื—ื•ืงืช ื‘ื™ืช ื”ื“ื™ืŸ, ื”ื—ืœื˜ื•ืชื™ื• ืฉืœ ื‘ื™ืช ื”ื“ื™ืŸ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ืื™ื ืŸ ื™ื•ืฆืจื•ืช ืชืงื“ื™ื ืžื—ื™ื™ื‘. ื‘ื›ืš, ื”ื—ืœื˜ื•ืชื™ื• ืžื—ื™ื™ื‘ื•ืช ืืช ื”ืžื“ื™ื ื•ืช ื”ืจืœื•ื•ื ื˜ื™ื•ืช ืœื“ื™ื•ืŸ, ื•ืœื ืืช ืืœื• ืฉืื™ื ืŸ. ืฉืœื™ืฉื™ืช, ื›ืชื‘ื™ื”ื ืฉืœ ื’ื“ื•ืœื™ ื”ืกื•ืคืจื™ื ื•ืžื•ืžื—ื™ ื”ืžืฉืคื˜ ืื™ื ื ืžื—ื™ื™ื‘ื™ื; ืืœื, ืœืคื™ ืก' 38(ื“) ืœื—ื•ืงืช ื‘ื™ืช ื”ื“ื™ืŸ - ืžืฉืžืฉื™ื ื›ืืžืฆืขื™ ืขื–ืจ, ื”ืžื”ื•ื•ื™ื ื›ืœื™ ืคืจืฉื ื™ ืœื”ื‘ื ืช ื”ืžืงื•ืจื•ืช ื”ืจืืฉื•ื ื™ื™ื. ืคื™ืจื•ืฉ ืฉืœ ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื›ืืฉืจ ื™ืฉ ืกื›ืกื•ื›ื™ื ืœื’ื‘ื™ ื”ืคื™ืจื•ืฉ ื”ืžื“ื•ื™ืง ื•ื”ื™ื™ืฉื•ื ืฉืœ ืžืฉืคื˜ ื‘ืชื•ืš ืžื“ื™ื ื”, ื”ืื—ืจื™ื•ืช ื”ื™ื ืฉืœ ื‘ืชื™ ื”ื“ื™ืŸ ืœืคืจืฉ ืืช ื”ืžืฉืคื˜. ื‘ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื›ื›ืœืœ, ืื™ืŸ ื‘ืชื™ ื“ื™ืŸ ื‘ืขืœื™ ืกืžื›ื•ืช ืžืกืคืงืช ืœื›ืš, ื•ื”ืื—ืจื™ื•ืช ืขืœ ืคื™ืจื•ืฉ ื”ืžืฉืคื˜ ื”ื™ื ื‘ื“ืจืš ื›ืœืœ ืฉืœ ื”ืžื“ื™ื ื•ืช ืขืฆืžืŸ. ืืžื ืช ื•ื™ื ื” ื‘ื“ื‘ืจ ื“ื™ื ื™ ืืžื ื•ืช ืงื•ื‘ืขืช "ืืžื ื” ืฆืจื™ื›ื” ืœื”ืชืคืจืฉ ื‘ื™ื•ืฉืจ, ืขืœ ืคื™ ื”ืžืฉืžืขื•ืช ื”ืจื’ื™ืœื” ืฉืœ ืžื•ื ื—ื™ ื”ื”ืกื›ื ื‘ื”ืงืฉืจื, ื•ืœืื•ืจ ืžื˜ืจืชื" (ืก' 31). ืืžื™ืจื” ื–ื• ื”ื™ื ื‘ืขืฆื ืคืฉืจื” ื‘ื™ืŸ ืฉืœื•ืฉ ืชืื•ืจื™ื•ืช ืฉื•ื ื•ืช ืฉืœ ืคื™ืจื•ืฉ: ืืœื” ื”ื ื”ื—ื•ืงื™ื ื”ื›ืœืœื™ื™ื ืฉืœ ื”ืคื™ืจื•ืฉ; ื—ื•ืงื™ื ืคืจื˜ื ื™ื™ื ืขืฉื•ื™ื™ื ืœื—ื•ืœ ื‘ืชื—ื•ืžื™ื ืฉื•ื ื™ื ืฉืœ ืžืฉืคื˜ ื‘ื™ืŸ-ืœืื•ืžื™. ืื›ื™ืคื” ืžืขื‘ืจ ืœื ื˜ื™ื™ื” ื”ื˜ื‘ืขื™ืช ืฉืœ ื”ืžื“ื™ื ื•ืช ืœืื›ื•ืฃ ื ื•ืจืžื•ืช ืžืกื•ื™ืžื•ืช, ื”ื›ื•ื— ืฉืœ ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ืชืžื™ื“ ื”ื’ื™ืข ืžื”ืœื—ืฅ ืฉื”ืžื“ื™ื ื•ืช ืžืคืขื™ืœื•ืช ืื—ืช ืขืœ ื”ืฉื ื™ื™ื” ืœื”ืชื ื”ื’ ื‘ืื•ืคืŸ ืขืงื™ื‘ ื•ืœื›ื‘ื“ ืืช ืžื—ื•ื™ื‘ื•ืชื™ื”ืŸ. ื›ืžื• ื‘ื›ืœ ืžืขืจื›ืช ืžืฉืคื˜ื™ืช, ืžืขืœื™ืžื™ื ืขื™ืŸ ืžื”ืคืจื•ืช ืจื‘ื•ืช ืฉืœ ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™. ืœืจื•ื‘, ื”ืฆื•ืจื” ื”ื™ื—ื™ื“ื” ืฉื‘ื” ืื•ื›ืคื™ื ืืช ื”ื—ื•ืง ื”ื•ื ื‘ืฆื™ื ื•ืจื•ืช ื“ื™ืคืœื•ืžื˜ื™ื™ื ื•ืขืœ ื™ื“ื™ ื”ืฉื ื”ืจืข ืฉื™ื•ืฆื ืœืžื“ื™ื ื” ืฉืขื•ื‘ืจืช ืขืœ ื”ื—ื•ืง. ืืฃ ืขืœ ืคื™ ืฉื”ื”ืคืจื•ืช ื”ืŸ ืœืžืขืฉื” ืจื‘ื•ืช, ืžื“ื™ื ื•ืช ืžื ืกื•ืช ืœื”ื™ืžื ืข ืžืœื”ื™ืจืื•ืช ื›ืขื•ื‘ืจื•ืช ืขืœ ื”ื—ื•ืง ื”ื‘ื™ืŸ-ืœืื•ืžื™. ืžื“ื™ื ื•ืช ื’ื ื™ื›ื•ืœื•ืช ืœื”ืคืขื™ืœ ืกื ืงืฆื™ื•ืช ื‘ืื•ืคืŸ ื—ื“-ืฆื“ื“ื™ ืื—ืช ืขืœ ื”ืฉื ื™ื™ื”, ื›ื’ื•ืŸ ื ื™ืชื•ืง ืงืฉืจื™ื ื“ื™ืคืœื•ืžื˜ื™ื™ื ืื• ื›ืœื›ืœื™ื™ื, ืื• ื‘ืคืขื•ืœื•ืช ื’ื•ืžืœื™ืŸ. ื‘ืžืงืจื™ื ืžืกื•ื™ืžื™ื, ื‘ืชื™ ื“ื™ืŸ ืื–ืจื—ื™ื™ื ื™ื›ื•ืœื™ื ืœืคืกื•ืง ื›ื ื’ื“ ืžื“ื™ื ื” ื–ืจื” (ื•ื–ื”ื• ื›ื‘ืจ ืžืขื ื™ื™ื ื• ืฉืœ ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ื”ืคืจื˜ื™) ื‘ืฉืœ ืคื’ื™ืขื”, ืืฃ ืฉื›ืืŸ ืžื“ื•ื‘ืจ ื‘ืขื ื™ื™ื ื™ื ืžืกื•ื‘ื›ื™ื, ืฉื‘ื”ื ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ืžืฆื˜ืœื‘ ืขื ื”ืื–ืจื—ื™. ืœืžื“ื™ื ื•ืช ื™ืฉื ื” ื”ื–ื›ื•ืช ืœื”ืฉืชืžืฉ ื‘ื›ื•ื— ืœื”ื’ื ื” ืขืฆืžื™ืช ื ื’ื“ ืžื“ื™ื ื” ืฉืคืขืœื” ื‘ื›ื•ื— ื ื’ื“ ื”ื˜ืจื™ื˜ื•ืจื™ื” ืฉืœื” ืื• ื”ืขืฆืžืื•ืช ื”ืคื•ืœื™ื˜ื™ืช ืฉืœื”. ืžื“ื™ื ื•ืช ื’ื ื™ื›ื•ืœื•ืช ืœื”ืฉืชืžืฉ ื‘ื›ื•ื— ืœื”ื’ื ื” ืขืฆืžื™ืช ืงื•ืœืงื˜ื™ื‘ื™ืช, ื›ืืฉืจ ื”ื›ื•ื— ืžื•ืคืขืœ ื ื’ื“ ืžื“ื™ื ื” ืื—ืจืช. ื”ืžื“ื™ื ื” ืฉื”ื•ืชืงืคื” ืฆืจื™ื›ื” ืœืืฉืจ ืืช ื”ื”ืฉืชืชืคื•ืช ืฉืœ ืžื“ื™ื ื•ืช ืื—ืจื•ืช ืœื”ื’ื ืชื” ื”ืขืฆืžื™ืช. ื–ื›ื•ืช ื–ื• ืžืขื•ื’ื ืช ื‘ืฆ'ืจื˜ืจ ืฉืœ ื”ืื•ืžื•ืช ื”ืžืื•ื—ื“ื•ืช. ื”ืคืจื•ืช ืฉืœ ืžื’ื™ืœืช ื”ืื•ืžื•ืช ื”ืžืื•ื—ื“ื•ืช (ื”ืื•"ื) ืขืœ ื™ื“ื™ ื—ื‘ืจื™ ื”ืื•"ื ื™ื›ื•ืœื•ืช ืœื”ื™ื“ื•ืŸ ื‘ืขืฆืจืช ื”ื›ืœืœื™ืช. ื”ืขืฆืจืช ื”ื›ืœืœื™ืช ืื™ื ื” ื™ื›ื•ืœื” ืœืงื‘ืœ ื”ื—ืœื˜ื•ืช ืคืขื•ืœื”, ืืš ืชื—ืช ื”ื”ื—ืœื˜ื” "ืžืชืื—ื“ื™ื ืœืฉืœื•ื" (Uniting for Peace, GA/RES/0377), ื”ื™ื ื”ืฆื”ื™ืจื” ื›ื™ ื”ื™ื ื™ื›ื•ืœื” ืœืืฉืจ ืฉื™ืžื•ืฉ ื‘ื›ื•ื— ืื ื”ื™ื™ืชื” ื”ืคืจื” ืฉืœ ื”ืฉืœื•ื ืื• ืคืขื•ืœื•ืช ืื™ื‘ื”, ืื ืžื•ืขืฆืช ื”ื‘ื™ื˜ื—ื•ืŸ ืœื ืคืขืœื” ื‘ืฉืœ ื”ืฆื‘ืขื” ืฉืœื™ืœื™ืช ืฉืœ ืื—ื“ ืžื”ื—ื‘ืจื™ื ื”ืงื‘ื•ืขื™ื (ื›ืœื•ืžืจ, ื•ื˜ื•). ื”ื™ื ื™ื›ื•ืœื” ืœืชื‘ื•ืข ืคืขื•ืœื•ืช ื›ืœืœื™ื•ืช ืื—ืจื•ืช (ื›ื’ื•ืŸ ืกื ืงืฆื™ื•ืช ืคื•ืœื™ื˜ื™ื•ืช) ืื ื”ืžืฆื‘ "ืื™ื•ื ืขืœ ื”ืฉืœื•ื" ืงื˜ืŸ ื™ื•ืชืจ. ื”ืžืฉืžืขื•ืช ื”ืžืฉืคื˜ื™ืช ืฉืœ ื”ื”ื—ืœื˜ื” ื”ื–ื• ืื™ื ื” ื‘ืจื•ืจื”, ืฉื›ืŸ ื”ืžื•ืขืฆื” ื”ื›ืœืœื™ืช ืื™ื ื” ื™ื›ื•ืœื” ืœื”ื—ืœื™ื˜ ื”ื—ืœื˜ื•ืช ืœืคืขื•ืœื”. ื ื™ืชืŸ ืœื”ืขืœื•ืช ื”ืคืจื•ืช ืœื“ื™ื•ืŸ ื’ื ื‘ืžื•ืขืฆืช ื”ื‘ื™ื˜ื—ื•ืŸ. ืžื•ืขืฆืช ื”ื‘ื™ื˜ื—ื•ืŸ ื™ื›ื•ืœื” ืœื”ื—ืœื™ื˜ ื”ื—ืœื˜ื•ืช ืชื—ืช ืคืจืง 6 ืฉืœ ืžื’ื™ืœืช ื”ืื•"ื, ื•ืœื”ืžืœื™ืฅ ืขืœ "ืคืชืจื•ืŸ ืœืกื›ืกื•ืš". ื”ื—ืœื˜ื•ืช ื›ืืœื” ืื™ื ืŸ ืงื•ื‘ืขื•ืช ืžื‘ื—ื™ื ืช ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™, ืื ื›ื™ ื”ื ื‘ื“ืจืš ื›ืœืœ ืžื‘ื˜ืื•ืช ืืช ื“ืขืช ื”ืžื•ืขืฆื”. ื‘ืžืงืจื™ื ื™ื•ืฆืื™ ื“ื•ืคืŸ, ืžื•ืขืฆืช ื”ื‘ื™ื˜ื—ื•ืŸ ื™ื›ื•ืœื” ืœื”ืขื‘ื™ืจ ื”ื—ืœื˜ื” ืชื—ืช ืคืจืง 7 ืฉืœ ื”ืžื’ื™ืœื”, ื”ืžืชื™ื™ื—ืก ืœ"ืื™ื•ืžื™ื ืขืœ ื”ืฉืœื•ื, ื”ืคืจื•ืช ืฉืœ ื”ืฉืœื•ื ื•ืคืขื•ืœื•ืช ืื™ื‘ื”", ื•ื”ื—ืœื˜ื•ืช ืืœื” ืงื•ื‘ืขื•ืช ืชื—ืช ื”ืžืฉืคื˜ ื”ื‘ื™ืŸ-ืœืื•ืžื™, ื•ื ื™ืชืŸ ืœื™ื™ืฉื ืื•ืชื ื‘ืกื ืงืฆื™ื•ืช ื›ืœื›ืœื™ื•ืช, ืคืขื•ืœื•ืช ืฆื‘ืื™ื•ืช, ืื• ืฉื™ืžื•ืฉ ืื—ืจ ื‘ื›ื•ื— ืชื—ืช ื”ื—ืกื•ืช ืฉืœ ื”ืื•"ื. ื ื˜ืขืŸ ื›ื™ ื’ื ื”ื—ืœื˜ื•ืช ืฉืœื ื”ื•ืขื‘ืจื• ืชื—ืช ืคืจืง 7 ื™ื›ื•ืœื•ืช ืœื”ื™ื•ืช ืงื•ื‘ืขื•ืช, ื•ื”ื‘ืกื™ืก ื”ืžืฉืคื˜ื™ ืœื–ื” ื”ื•ื ื”ื›ื•ื—ื•ืช ื”ืจื—ื‘ื™ื ื™ื•ืชืจ ืฉืœ ื”ืžื•ืขืฆื” ืชื—ืช ืกืขื™ืฃ 24(2), ืฉืงื•ื‘ืข ื›ื™ "ื‘ื‘ื™ืฆื•ืข ื—ื•ื‘ื•ืช ืืœื” (ืื—ืจื™ื•ืช ืขืœ ื”ืฉืœื•ื ื•ื”ื‘ื™ื˜ื—ื•ืŸ ื”ื‘ื™ืŸ-ืœืื•ืžื™) ื”ืžื•ืขืฆื” ืชืคืขืœ ื‘ืชื™ืื•ื ืขื ื”ืžื˜ืจื•ืช ื•ื”ืขืงืจื•ื ื•ืช ืฉืœ ื”ืื•ืžื•ืช ื”ืžืื•ื—ื“ื•ืช". ื”ืื•ืคื™ ื”ืงื•ื‘ืข ืฉืœ ื”ื—ืœื˜ื•ืช ื›ืืœื” ื ืชืžืš ื‘ื™ื“ื™ ื‘ื™ืช ื”ื“ื™ืŸ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ืœืฆื“ืง ื‘ื“ืขื” ื”ืžื™ื™ืขืฆืช ืฉืœื• ืขืœ ื ืžื™ื‘ื™ื”. ืžื“ื™ื ื•ืช ื™ื›ื•ืœื•ืช ื’ื, ื‘ื”ืกื›ืžื” ื”ื“ื“ื™ืช, ืœื”ื‘ื™ื ืกื›ืกื•ื›ื™ื ืœื’ื™ืฉื•ืจ ื•ืคืฉืจื” ืœื‘ื™ืช ื”ื“ื™ืŸ ื”ื‘ื™ืŸ-ืœืื•ืžื™ ืœืฆื“ืง, ืฉืžืžื•ืงื ื‘ื”ืื’, ื‘ื”ื•ืœื ื“. ื”ืคืกืงื™ื ืฉื ื™ืชื ื™ื ืขืœ ื™ื“ื™ ื‘ื™ืช ื”ื“ื™ืŸ ื‘ืžืงืจื™ื ื›ืืœื” ื”ื ืงื•ื‘ืขื™ื, ืืฃ ื›ื™ ืื™ืŸ ืœื• ื›ืœื™ื ืœืื›ื•ืฃ ืืช ื”ื—ืœื˜ื•ืชื™ื•. ื‘ื™ืช ื”ื“ื™ืŸ ื™ื›ื•ืœ ืœืชืช ื“ืขื” ืžื™ื™ืขืฆืช ืขืœ ื›ืœ ืฉืืœื” ืžืฉืคื˜ื™ืช ืœื‘ืงืฉืช ื›ืœ ื’ื•ืฃ ืฉืงื™ื‘ืœ ืืช ื”ืกืžื›ื•ืช ืœื‘ืงืฉื” ื›ื–ื• ืžื”ืฆ'ืจื˜ืจ ืฉืœ ื”ืื•"ื. ื‘ื›ืžื” ืžืžืงืจื™ ื”ื™ื™ืขื•ืฅ ืฉื”ื’ื™ืขื• ืืœ ื‘ื™ืช ื”ื“ื™ืŸ, ื”ื™ื• ืžื—ืœื•ืงื•ืช ืœื’ื‘ื™ ืžื™ื“ืช ื”ืกืžื›ื•ืช ื•ื”ื™ื›ื•ืœืช ืฉืœ ื‘ื™ืช ื”ื“ื™ืŸ. ืžืงืจื™ื ืฉืžื’ื™ืขื™ื ืœื‘ื™ืช ื”ื“ื™ืŸ ื”ื ืคืขืžื™ื ืจื‘ื•ืช ืžืกื•ื‘ื›ื™ื ืžืื•ื“ (ื”ื™ื• ืจืง 150 ืžืงืจื™ื ื›ืืœื” ืžืื– ื ื•ืกื“ ื‘ื™ืช ื”ื“ื™ืŸ ื‘-1945), ื•ื”ื ื™ื›ื•ืœื™ื ืœื”ื™ืžืฉืš ืฉื ื™ื ืืจื•ื›ื•ืช ื•ืœืžืœื ืืœืคื™ ื“ืคื™ื ืฉืœ ืจืื™ื•ืช ื•ืขืจืขื•ืจื™ื, ื•ืœื”ืขืกื™ืง ืืช ืขื•ืจื›ื™ ื”ื“ื™ืŸ ื”ื˜ื•ื‘ื™ื ื‘ื™ื•ืชืจ ืœืžืฉืคื˜ ืฆื™ื‘ื•ืจื™ ื‘ื™ืŸ-ืœืื•ืžื™ ื‘ืขื•ืœื. ื‘ืฉื ืช 2005, ื”ื™ื• 12 ืžืงืจื™ื ื‘ื“ื™ื•ืŸ ืœืคื ื™ ื‘ื™ืช ื”ื“ื™ืŸ. ื”ื—ืœื˜ื•ืช ื‘ืžืงืจื™ื ืฉืœ ื’ื™ืฉื•ืจ ืขืฉื•ื™ื•ืช ืœื”ื™ื•ืช ืงื•ื‘ืขื•ืช ืื• ืœื, ืขืœ ืคื™ ื”ืกื›ื ื”ื’ื™ืฉื•ืจ, ื•ืื™ืœื• ื”ื—ืœื˜ื•ืช ื‘ืžืงืจื™ื ืฉืœ ืกื›ืกื•ืš ืœืœื ื’ื™ืฉื•ืจ ืฉื ื˜ืขื ื™ื ื‘ืคื ื™ ื‘ื™ืช ื”ื“ื™ืŸ, ื”ืŸ ืชืžื™ื“ ืงื•ื‘ืขื•ืช ื›ืœืคื™ ื”ืžื“ื™ื ื•ืช ื”ืžืขื•ืจื‘ื•ืช. ืืฃ ืขืœ ืคื™ ืฉืžื“ื™ื ื•ืช (ื•ื‘ื–ืžืŸ ื”ืื—ืจื•ืŸ, ื’ื ืืจื’ื•ื ื™ื ื‘ื™ืŸ-ืœืื•ืžื™ื™ื) ื”ื ื‘ื“ืจืš ื›ืœืœ ื”ื™ื—ื™ื“ื•ืช ืฉืคื•ืขืœื•ืช ื›ืฆื“ื“ื™ื ื‘ืžืฉืคื˜ ื‘ื™ืŸ-ืœืื•ืžื™, ืœืืžื ื•ืช ืžืกื•ื™ืžื•ืช, ื›ื’ื•ืŸ ื”ืืžื ื” ื”ื‘ื™ืŸ-ืœืื•ืžื™ืช ืœื–ื›ื•ื™ื•ืช ืื–ืจื—ื™ื•ืช ื•ืคื•ืœื™ื˜ื™ื•ืช, ื™ืฉ ืคืจื•ื˜ื•ืงื•ืœ ืื•ืคืฆื™ื•ื ืœื™ ื”ืžืืคืฉืจ ืœื™ื—ื™ื“ื™ื ืฉื–ื›ื•ื™ื•ืชื™ื”ื ื”ื•ืคืจื• ื‘ื™ื“ื™ ืžื“ื™ื ื•ืช ื—ื‘ืจื•ืช, ืœืขืชื•ืจ ืœื•ื•ืขื™ื“ืช ื–ื›ื•ื™ื•ืช ื”ืื“ื ื”ื‘ื™ืŸ-ืœืื•ืžื™ืช. ืœืขืจื›ื™ื, ืชื—ื•ืžื™ื ื•ืžื•ืฉื’ื™ื ื ื•ืกืคื™ื ืจืื• ืคื•ืจื˜ืœ ื—ื•ืง ื•ืžืฉืคื˜ ืจืื• ื’ื ืœืงืจื™ืื” ื ื•ืกืคืช ืงื™ืฉื•ืจื™ื ื—ื™ืฆื•ื ื™ื™ื ื”ืขืจื•ืช ืฉื•ืœื™ื™ื ื”ื‘ื”ืจื”: ื”ืžื™ื“ืข ื‘ื•ื•ื™ืงื™ืคื“ื™ื” ื ื•ืขื“ ืœื”ืขืฉืจื” ื‘ืœื‘ื“ ื•ืื™ืŸ ืœืจืื•ืช ื‘ื• ื™ื™ืขื•ืฅ ืžืฉืคื˜ื™.
========================================
[SOURCE: https://www.lomdimbareshet.net/%d7%94%d7%9b%d7%a0%d7%94-%d7%9c%d7%aa%d7%99%d7%9b%d7%95%d7%9f-3-%d7%99%d7%97%d7%99%d7%93%d7%95%d7%aa/] | [TOKENS: 529]
ื”ื›ื ื” ืœืชื™ื›ื•ืŸ - 3 ื™ื—ื™ื“ื•ืช ืžื™ื™ืฉืจื™ื ืงื• ื”ื™ื ื—ื•ื‘ืจืช ื”ื›ื ื” ืœืชื™ื›ื•ืŸ ื”ื›ื•ืœืœืช ื ื•ืฉืื™ื ื”ื ืœืžื“ื™ื ื‘ืชื™ื›ื•ืŸ ื‘ืจืžืช 3 ื™ื—ื™ื“ื•ืช, ื•ืžื—ื•ืœืงืช ืœืคืจืง ืืœื’ื‘ืจื” ื•ืคืจืง ื’ื™ืื•ืžื˜ืจื™ื”.ื”ื—ื•ื‘ืจืช ืžื•ืžืœืฆืช ืœืœืžื™ื“ื” ืœืคื ื™ ื”ืžืขื‘ืจ ืœื›ื™ืชื” ื™' . ืœืžื™ื“ื” ื‘ืขื–ืจืช ื—ื•ื‘ืจืช ืœื™ืžื•ื“ ืžื’ื“ื™ืœื” ืืช ื”ื”ื‘ื ื” ื•ื—ื•ืฉืคืช ืืช ื”ืœื•ืžื“ ืœื”ืกื‘ืจื™ื ื•ื“ื•ื’ืžืื•ืช ืžืคื•ืจื˜ื•ืช ืœื›ืœ ืื—ื“ ืžื ื•ืฉืื™ ื”ืงื•ืจืก. ื ื™ืชืŸ ืœื”ื“ืคื™ืก ืืช ื”ื—ื•ื‘ืจืช ื•ืœืชืจื’ืœ ื‘ืืžืฆืขื•ืชื” ืฉืืœื•ืช ืจื‘ื•ืช ื‘ื›ืœ ื ื•ืฉื. ื”ื—ื•ื‘ืจืช ืžื—ื•ืœืงืช ืœืคืจืงื™ ืœื™ืžื•ื“ ื”ื ื™ืชื ื™ื ืœืœืžื™ื“ื” ื‘ืฆื•ืจื” ืขืฆืžืื™ืช ืžืจืžืช ื”ื‘ืกื™ืก ื•ืขื“ ืœืจืžืช ื”ื‘ื’ืจื•ืช. ืœื”ืชืจืฉืžื•ืช โ€“ ืœื—ืฆื• ืขืœ ื”ืชืžื•ื ื”. ืฆืจื• ืงืฉืจ - ื ืฉืžื— ืœืขื–ื•ืจ. ยฉ 2024 ื›ืœ ื”ื–ื›ื•ื™ื•ืช ืฉืžื•ืจื•ืช. ืœื•ืžื“ื™ื ื‘ืจืฉืช - ื“ื•ื“ื• ื’ื•ืœื“ืฉื˜ื™ื™ืŸ ืขืจื•ืฅ "ืœื•ืžื“ื™ื ื‘ืจืฉืช" ืœื™ืžื•ื“ ืขืฆืžื™
========================================
[SOURCE: https://mschool.co.il/reg2026school/] | [TOKENS: 435]
ื”ืจืฉืžื” ืœืžืขืจื›ืช ื”ืจืฉืžื” ืœืืชืจ ื”ืงื•ืจืกื™ื ืฉื™ืžื• ืœื‘! ื™ืฉ ืœื”ื™ืจืฉื ืขื ืื™ืžื™ื™ืœ ืขื“ื›ื ื™.ืฉื ืžืฉืชืžืฉ ื‘ืื ื’ืœื™ืช ื‘ืœื‘ื“ (ืืคืฉืจื™ ืฉื™ืœื•ื‘ ืื ื’ืœื™ืช ื•ืžืกืคืจื™ื) ืœืœื ืกื™ืžื ื™ื ืžื™ื•ื—ื“ื™ื ื•ืœืœื ืจื•ื•ื—.ืœื“ื•ื’ืžื daniel2056 ื”ืจืฉืžื” ืจืืฉื•ื ื™ืช ืœืžืขืจื›ืช ื›ืชื•ื‘ืช ืื™ืžื™ื™ืœ * ืฉื ืžืฉืชืžืฉ (ื‘ืื ื’ืœื™ืช ื‘ืœื‘ื“ ื•ืžืกืคืจื™ื - ืœืœื ืจื•ื•ื— ื•ืกื™ืžื ื™ื ืžื™ื•ื—ื“ื™ื) * ืกื™ืกืžื * ืื™ืฉื•ืจ ื”ืกื™ืกืžื ื”ื—ื“ืฉื” * ื”ืจืฉืžื” ื”ืจืฉืžื” ืœืืชืจ ื”ืงื•ืจืกื™ื ืฉื™ืžื• ืœื‘! ื™ืฉ ืœื”ื™ืจืฉื ืขื ืื™ืžื™ื™ืœ ืขื“ื›ื ื™. ืฉื ืžืฉืชืžืฉ ื‘ืื ื’ืœื™ืช ื‘ืœื‘ื“ (ืืคืฉืจื™ ืฉื™ืœื•ื‘ ืื ื’ืœื™ืช ื•ืžืกืคืจื™ื) ืœืœื ืกื™ืžื ื™ื ืžื™ื•ื—ื“ื™ื ื•ืœืœื ืจื•ื•ื—. ืœื“ื•ื’ืžื daniel2056 ื”ืจืฉืžื” ืชืคืจื™ื˜ ื ื’ื™ืฉื•ืช
========================================
[SOURCE: https://huggingface.co/datasets?modality=modality%3Atabular] | [TOKENS: 915]
Datasets OpenResearcher/OpenResearcher-Dataset Viewer โ€ข Updated 9 days ago โ€ข 97.6k โ€ข 8.47k โ€ข 72 allenai/molmospaces Viewer โ€ข Updated 5 days ago โ€ข 772k โ€ข 206 โ€ข 38 atreydesai/qgqa-gpt-5.2-20260213-041705 Viewer โ€ข Updated 8 days ago โ€ข 3k โ€ข 76 โ€ข 36 OpenDriveLab-org/Kai0 Viewer โ€ข Updated 7 days ago โ€ข 1.8k โ€ข 24.7k โ€ข 30 HuggingFaceFW/fineweb Viewer โ€ข Updated Jul 11, 2025 โ€ข 52.5B โ€ข 194k โ€ข 2.67k nyuuzyou/suno Viewer โ€ข Updated 18 days ago โ€ข 660k โ€ข 374 โ€ข 132 HuggingFaceFW/fineweb-edu Viewer โ€ข Updated Jul 11, 2025 โ€ข 3.5B โ€ข 254k โ€ข 954 openfoodfacts/product-database Viewer โ€ข Updated about 21 hours ago โ€ข 4.35M โ€ข 4.17k โ€ข 82 Idavidrein/gpqa Benchmark โ€ข Updated 30 days ago โ€ข 1.25k โ€ข 86.6k โ€ข 367 PleIAs/common_corpus Viewer โ€ข Updated 2 days ago โ€ข 69.9k โ€ข 60k โ€ข 354 lm-provers/FineProofs-SFT Viewer โ€ข Updated 7 days ago โ€ข 12.1k โ€ข 74 โ€ข 11 TIGER-Lab/MMLU-Pro Benchmark โ€ข Updated Jan 19 โ€ข 12.1k โ€ข 83.4k โ€ข 431 MathArena/aime_2026 Benchmark โ€ข Updated 5 days ago โ€ข 30 โ€ข 538 โ€ข 10 nvidia/PhysicalAI-Robotics-GR00T-Teleop-GR1 Viewer โ€ข Updated 7 days ago โ€ข 7.55M โ€ข 1.09k โ€ข 8 bigcode/the-stack-v2 Viewer โ€ข Updated Apr 23, 2024 โ€ข 5.45B โ€ข 7.32k โ€ข 472 HuggingFaceTB/smollm-corpus Viewer โ€ข Updated Sep 6, 2024 โ€ข 237M โ€ข 22.1k โ€ข 437 uw-math-ai/theorem-search-dataset Viewer โ€ข Updated 1 day ago โ€ข 2.89M โ€ข 240 โ€ข 21 ronantakizawa/leetcode-assembly Viewer โ€ข Updated 5 days ago โ€ข 14.1k โ€ข 88 โ€ข 7 kmfoda/booksum Viewer โ€ข Updated Nov 30, 2022 โ€ข 12.5k โ€ข 1.31k โ€ข 71 bowen-upenn/PersonaMem-v2 Viewer โ€ข Updated 16 days ago โ€ข 51.7k โ€ข 2.31k โ€ข 19 cx-cmu/deepresearchgym-agentic-search-logs Viewer โ€ข Updated 22 days ago โ€ข 14.3M โ€ข 113 โ€ข 12 allenai/real-toxicity-prompts Viewer โ€ข Updated Sep 30, 2022 โ€ข 99.4k โ€ข 7.4k โ€ข 113 jtatman/stable-diffusion-prompts-stats-full-uncensored Viewer โ€ข Updated Nov 8, 2024 โ€ข 897k โ€ข 359 โ€ข 120 HuggingFaceFW/fineweb-2 Viewer โ€ข Updated Oct 27, 2025 โ€ข 4.48B โ€ข 72.4k โ€ข 755 HHS-Official/medicaid-provider-spending Viewer โ€ข Updated 4 days ago โ€ข 227M โ€ข 171 โ€ข 4 seongsubae/KorMedMCQA-V Viewer โ€ข Updated 4 days ago โ€ข 1.84k โ€ข 518 โ€ข 4 deepmind/code_contests Viewer โ€ข Updated Jun 11, 2023 โ€ข 4.04k โ€ข 861k โ€ข 216 OpenAssistant/oasst1 Viewer โ€ข Updated May 2, 2023 โ€ข 88.8k โ€ข 11.2k โ€ข 1.48k HuggingFaceTB/smoltalk Viewer โ€ข Updated Feb 10, 2025 โ€ข 2.2M โ€ข 5.75k โ€ข 391 openbmb/DCAD-2000 Viewer โ€ข Updated Dec 5, 2025 โ€ข 649M โ€ข 3.99k โ€ข 20
========================================
[SOURCE: https://huggingface.co/datasets/Raziel1234/WebBooks/discussions/1#6918ba7c06ef9e60dba2f08a] | [TOKENS: 343]
Datasets: Raziel1234 / WebBooks like 0 [bot] Conversion to Parquet The parquet-converter bot has created a version of this dataset in the Parquet format in the refs/convert/parquet branch. What is Parquet? Apache Parquet is a popular columnar storage format known for: This is what powers the dataset viewer on each dataset page and every dataset on the Hub can be accessed with the same code (you can use HF Datasets, ClickHouse, DuckDB, Pandas, PostgreSQL, or Polars, up to you). You can learn more about the advantages associated with Parquet in the documentation. How to access the Parquet version of the dataset? You can access the Parquet version of the dataset by following this link: refs/convert/parquet What if my dataset was already in Parquet? When the dataset is already in Parquet format, the data are not converted and the files in refs/convert/parquet are links to the original files. This rule has an exception to ensure the dataset viewer API to stay fast: if the row group size of the original Parquet files is too big, new Parquet files are generated. What should I do? You don't need to do anything. The Parquet version of the dataset is available for you to use. Refer to the documentation for examples and code snippets on how to query the Parquet files with ClickHouse, DuckDB, Pandas or Polars. If you have any questions or concerns, feel free to ask in the discussion below. You can also close the discussion if you don't have any questions. ืฉืœื•ื ยท Sign up or log in to comment
========================================
[SOURCE: https://huggingface.co/models?pipeline_tag=text-to-speech] | [TOKENS: 987]
Models Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice Text-to-Speech โ€ข Updated 23 days ago โ€ข 933k โ€ข 1.12k nineninesix/kani-tts-2-en Text-to-Speech โ€ข 0.4B โ€ข Updated 2 days ago โ€ข 2.59k โ€ข 163 OpenMOSS-Team/MOSS-TTS Text-to-Speech โ€ข 8B โ€ข Updated 8 days ago โ€ข 41.2k โ€ข 289 Soul-AILab/SoulX-Singer Text-to-Speech โ€ข Updated 10 days ago โ€ข 1.08k โ€ข 125 hexgrad/Kokoro-82M Text-to-Speech โ€ข Updated Apr 10, 2025 โ€ข 8.66M โ€ข โ€ข 5.73k Aratako/MioTTS-2.6B Text-to-Speech โ€ข Updated 11 days ago โ€ข 924 โ€ข 63 nineninesix/kani-tts-2-pt Text-to-Speech โ€ข 0.4B โ€ข Updated 2 days ago โ€ข 1.19k โ€ข 37 NAMAA-Space/NAMAA-Saudi-TTS Text-to-Speech โ€ข 0.5B โ€ข Updated 23 days ago โ€ข 138 โ€ข 35 OpenMOSS-Team/MOSS-TTS-Realtime Text-to-Speech โ€ข 2B โ€ข Updated 6 days ago โ€ข 10.4k โ€ข 50 OpenMOSS-Team/MOSS-TTSD-v1.0 Text-to-Speech โ€ข 8B โ€ข Updated 7 days ago โ€ข 11.9k โ€ข 44 Qwen/Qwen3-TTS-12Hz-1.7B-VoiceDesign Text-to-Speech โ€ข 2B โ€ข Updated 23 days ago โ€ข 406k โ€ข 266 microsoft/VibeVoice-Realtime-0.5B Text-to-Speech โ€ข 1B โ€ข Updated Dec 12, 2025 โ€ข 602k โ€ข 1.12k YatharthS/LuxTTS Text-to-Speech โ€ข Updated 29 days ago โ€ข 2.81k โ€ข 138 OpenMOSS-Team/MOSS-VoiceGenerator Text-to-Speech โ€ข 2B โ€ข Updated 10 days ago โ€ข 4.66k โ€ข 34 coqui/XTTS-v2 Text-to-Speech โ€ข Updated Dec 11, 2023 โ€ข 7.19M โ€ข 3.4k ResembleAI/chatterbox Text-to-Speech โ€ข Updated Sep 23, 2025 โ€ข 766k โ€ข โ€ข 1.48k kugelaudio/kugelaudio-0-open Text-to-Speech โ€ข Updated 15 days ago โ€ข 91.4k โ€ข 163 OpenMOSS-Team/MOSS-TTS-Local-Transformer Text-to-Speech โ€ข 3B โ€ข Updated 8 days ago โ€ข 22.7k โ€ข 19 syvai/plapre-nano Text-to-Speech โ€ข 0.3B โ€ข Updated 3 days ago โ€ข 1.04k โ€ข 9 neuphonic/neutts-nano Text-to-Speech โ€ข 0.2B โ€ข Updated 9 days ago โ€ข 6.41k โ€ข 52 fishaudio/s1-mini Text-to-Speech โ€ข Updated 15 days ago โ€ข 5.21k โ€ข 590 Aratako/MioTTS-GGUF Text-to-Speech โ€ข 0.1B โ€ข Updated 12 days ago โ€ข 3.01k โ€ข 14 neuphonic/neutts-nano-q8-gguf Text-to-Speech โ€ข 0.2B โ€ข Updated 9 days ago โ€ข 2.02k โ€ข 11 bharatgenai/sooktam2 Text-to-Speech โ€ข Updated about 6 hours ago โ€ข 222 โ€ข 6 vadimbelsky/emirati-vits-male-1.0 Text-to-Speech โ€ข Updated 2 days ago โ€ข 6 fishaudio/fish-speech-1.5 Text-to-Speech โ€ข Updated Mar 25, 2025 โ€ข 3.82k โ€ข 709 sesame/csm-1b Text-to-Speech โ€ข Updated Dec 1, 2025 โ€ข 129k โ€ข 2.34k microsoft/VibeVoice-1.5B Text-to-Speech โ€ข 3B โ€ข Updated about 1 month ago โ€ข 77.3k โ€ข 2.22k zai-org/GLM-TTS Text-to-Speech โ€ข Updated Jan 12 โ€ข 250 โ€ข 320 FunAudioLLM/Fun-CosyVoice3-0.5B-2512 Text-to-Speech โ€ข Updated 18 days ago โ€ข 5.78k โ€ข 459
========================================
[SOURCE: https://huggingface.co/docs/hub/enterprise-hub-datasets] | [TOKENS: 105]
Hub documentation Datasets Hub and get access to the augmented documentation experience to get started Datasets This feature is part of the Team & Enterprise plans. Data Studio is enabled on private datasets under your Team or Enterprise organization. Data Studio allows teams to understand their data and to help them build better data processing and filtering for AI. This powerful viewer allows you to explore dataset content, inspect data distributions, filter by values, search for keywords, or even run SQL queries on your data without leaving your browser. More information about Data Studio.
========================================
[SOURCE: https://huggingface.co/models?other=chemistry] | [TOKENS: 803]
Models Prior-Labs/tabpfn_2_5 Tabular Classification โ€ข Updated 23 days ago โ€ข 95.4k โ€ข 162 thelamapi/next-1b Text Generation โ€ข 1.0B โ€ข Updated Nov 11, 2025 โ€ข 3.51k โ€ข 25 SandboxAQ/AQAffinity Updated 11 days ago โ€ข 23 VibeStudio/Nidum-Llama-3.2-3B-Uncensored-GGUF Text Generation โ€ข 3B โ€ข Updated Dec 17, 2024 โ€ข 3.89k โ€ข 35 bartowski/PocketDoc_Dans-PersonalityEngine-V1.3.0-24b-GGUF Text Generation โ€ข 24B โ€ข Updated May 23, 2025 โ€ข 29.1k โ€ข 35 microsoft/NatureLM-8x7B 47B โ€ข Updated Jun 20, 2025 โ€ข 155 โ€ข 19 OpenMed/OpenMed-NER-ChemicalDetect-ElectraMed-33M Token Classification โ€ข 33.2M โ€ข Updated Aug 5, 2025 โ€ข 191k โ€ข โ€ข 2 Moreza009/Llama-DrugReasoner Text Generation โ€ข Updated Oct 7, 2025 โ€ข 6 โ€ข 2 ByteDance-Seed/byteff2 Updated Nov 17, 2025 โ€ข 3 SandboxAQ/aqcat25-ev2 Updated Oct 30, 2025 โ€ข 9 thelamapi/next-ocr Image-Text-to-Text โ€ข 9B โ€ข Updated Nov 15, 2025 โ€ข 10k โ€ข 16 ValiantLabs/Ministral-3-14B-Reasoning-2512-ShiningValiant3 Text Generation โ€ข 14B โ€ข Updated Dec 8, 2025 โ€ข 93 โ€ข 4 vipsehgal/qwen3-8b-jee-sft Text Generation โ€ข 8B โ€ข Updated 3 days ago โ€ข 57 โ€ข 1 ngetichkpeter/Logoi Updated 6 days ago โ€ข 1 Felipe2231/canario-amarelo-prod-final Text-to-Video โ€ข Updated 7 days ago โ€ข 1 Intae/mymodel Updated Aug 24, 2023 davanstrien/test Updated Sep 7, 2023 โ€ข 13 seyonec/ChemBERTa-zinc-base-v1 Fill-Mask โ€ข Updated May 20, 2021 โ€ข 651k โ€ข โ€ข 62 ncfrey/ChemGPT-4.7M Text Generation โ€ข Updated Jun 15, 2022 โ€ข 1.15k โ€ข 20 ncfrey/ChemGPT-19M Text Generation โ€ข Updated Jun 15, 2022 โ€ข 660 โ€ข 4 ncfrey/ChemGPT-1.2B Text Generation โ€ข Updated Jun 15, 2022 โ€ข 779 โ€ข 16 sagawa/ReactionT5v1-forward Updated Jul 28, 2024 โ€ข 123 Dr-BERT/DrBERT-4GB Fill-Mask โ€ข Updated May 28, 2023 โ€ข 196 โ€ข 1 Dr-BERT/DrBERT-7GB Fill-Mask โ€ข Updated May 28, 2023 โ€ข 2.28k โ€ข โ€ข 17 UEG/interface Text Classification โ€ข Updated Feb 27, 2023 Sevenlee/kkk Image Segmentation โ€ข Updated Jan 10, 2023 Dr-BERT/DrBERT-4GB-CP-CamemBERT Updated May 28, 2023 Dr-BERT/DrBERT-4GB-CP-PubMedBERT Fill-Mask โ€ข Updated May 28, 2023 โ€ข 355 โ€ข 1 cosmobaby/ka Image Segmentation โ€ข Updated Jan 10, 2023 csimonmeunier/test-model Updated Jan 23, 2023 โ€ข 1
========================================
[SOURCE: https://huggingface.co/docs/hub/spaces-dev-mode] | [TOKENS: 724]
Hub documentation Spaces Dev Mode: Seamless development in Spaces Hub and get access to the augmented documentation experience to get started Spaces Dev Mode: Seamless development in Spaces This feature is still in Beta stage. The Spaces Dev Mode is part of PRO or Team & Enterprise plans. Spaces Dev Mode Spaces Dev Mode is a feature that eases the debugging of your application and makes iterating on Spaces faster. Whenever your commit some changes to your Space repo, the underlying Docker image gets rebuilt, and then a new virtual machine is provisioned to host the new container. The Dev Mode allows you to update your Space much quicker by overriding the Docker image. The Dev Mode Docker image starts your application as a sub-process, allowing you to restart it without stopping the Space container itself. It also starts a VS Code server and a SSH server in the background for you to connect to the Space. The ability to connect to the running Space unlocks several use cases: Overall it makes developing and experimenting with Spaces much faster by skipping the Docker image rebuild phase. Interface Once the Dev Mode is enabled on your Space, you should see a modal like the following. The application does not restart automatically when you change the code. For your changes to appear in the Space, you need to use the Refresh button that will restart the app. The Dev Mode allows you to connect to your Spaceโ€™s docker container using SSH. Instructions to connect are listed in the Dev Mode controls modal. You will need to add your machineโ€™s SSH public key to your user account to be able to connect to the Space using SSH. Check out the Git over SSH documentation for more detailed instructions. You can also use a local install of VS Code to connect to the Space container. To do so, you will need to install the SSH Remote extension. The modal will display a warning if you have uncommitted or unpushed changes in the Space: Enabling Dev Mode You can enable the Dev Mode on your Space from the web interface. You can also create a Space with the dev mode enabled: Limitations Dev Mode is currently not available for static Spaces. Docker Spaces also have some additional requirements. Dev Mode is supported for Docker Spaces. However, your Space needs to comply with the following rules for Dev Mode to work properly. Your application code must be located in the /app folder for the Dev Mode daemon to be able to detect changes. The /app folder must be owned by the user with uid 1000 to allow you to make changes to the code. The Dockerfile must contain a CMD instruction for startup. Checkout Dockerโ€™s documentation about the CMD instruction for more details. Dev Mode works well when the base image is debian-based (eg, ubuntu). More exotic linux distros (eg, alpine) are not tested and Dev Mode is not guaranteed to work on them. This is an example of a Dockerfile compatible with Spaces Dev Mode. It installs the required packages with apt-get, along with a couple more for developer convenience (namely: top, vim and nano). It then starts a NodeJS application from /app. There are several examples of Dev Mode compatible Docker Spaces in this organization. Feel free to duplicate them in your namespace! Example Python app (FastAPI HTTP server): https://huggingface.co/spaces/dev-mode-explorers/dev-mode-python Example Javascript app (Express.js HTTP server): https://huggingface.co/spaces/dev-mode-explorers/dev-mode-javascript Feedback You can share your feedback on Spaces Dev Mode directly on the HF Hub: https://huggingface.co/spaces/dev-mode-explorers/README/discussions
========================================
[SOURCE: https://huggingface.co/models?other=biology] | [TOKENS: 877]
Models BioMistral/BioMistral-7B-GGUF Text Generation โ€ข 7B โ€ข Updated Feb 19, 2024 โ€ข 962 โ€ข 19 Prior-Labs/tabpfn_2_5 Tabular Classification โ€ข Updated 23 days ago โ€ข 95.4k โ€ข 162 vandijklab/C2S-Scale-Gemma-2-27B Text Generation โ€ข 28B โ€ข Updated Oct 31, 2025 โ€ข 1.58k โ€ข 161 thelamapi/next-1b Text Generation โ€ข 1.0B โ€ข Updated Nov 11, 2025 โ€ข 3.51k โ€ข 25 SandboxAQ/AQAffinity Updated 11 days ago โ€ข 23 google/alphagenome-all-folds Updated 25 days ago โ€ข 76 google/alphagenome-fold-1 Updated 25 days ago โ€ข 8 thefynnbe/ambitious-sloth Updated 9 days ago โ€ข 2 imageomics/bioclip-2.5-vith14 Zero-Shot Image Classification โ€ข Updated 8 days ago โ€ข 87 โ€ข 2 zhihan1996/DNABERT-2-117M Updated Jun 30, 2025 โ€ข 60.7k โ€ข 89 Rostlab/ProstT5 Translation โ€ข Updated Nov 16, 2023 โ€ข 45.8k โ€ข 33 chriamue/bird-species-classifier Image Classification โ€ข 8.51M โ€ข Updated Nov 12, 2023 โ€ข 1.24k โ€ข โ€ข 21 TheBloke/medicine-chat-GGUF Text Generation โ€ข 7B โ€ข Updated Jan 10, 2024 โ€ข 410 โ€ข 17 BioMistral/BioMistral-7B Text Generation โ€ข Updated Feb 21, 2024 โ€ข 41.2k โ€ข 492 EvolutionaryScale/esm3-sm-open-v1 Updated Jan 29, 2025 โ€ข 11.3k โ€ข 277 biomap-research/proteinglm-100b-int4 50B โ€ข Updated Mar 17, 2025 โ€ข 188 โ€ข 11 minwoosun/uce-650m Updated Jul 24, 2024 โ€ข 3 โ€ข 3 minwoosun/uce-100m Updated Jul 24, 2024 โ€ข 28 โ€ข 2 vandijklab/C2S-Pythia-410m-cell-type-prediction Text Generation โ€ข 0.4B โ€ข Updated Oct 31, 2025 โ€ข 738 โ€ข 10 johahi/borzoi-replicate-0 0.2B โ€ข Updated Jan 3, 2025 โ€ข 224k โ€ข 1 genbio-ai/AIDO.Protein-16B Updated Nov 13, 2025 โ€ข 330 โ€ข 6 VibeStudio/Nidum-Llama-3.2-3B-Uncensored-GGUF Text Generation โ€ข 3B โ€ข Updated Dec 17, 2024 โ€ข 3.89k โ€ข 35 codewithdark/vit-chest-xray Image Classification โ€ข 85.8M โ€ข Updated Aug 26, 2025 โ€ข 1.17k โ€ข โ€ข 5 bartowski/PocketDoc_Dans-PersonalityEngine-V1.3.0-24b-GGUF Text Generation โ€ข 24B โ€ข Updated May 23, 2025 โ€ข 29.1k โ€ข 35 imageomics/bioclip-2 Zero-Shot Image Classification โ€ข Updated 8 days ago โ€ข 15.9k โ€ข 28 microsoft/NatureLM-8x7B 47B โ€ข Updated Jun 20, 2025 โ€ข 155 โ€ข 19 Moreza009/Llama-DrugReasoner Text Generation โ€ข Updated Oct 7, 2025 โ€ข 6 โ€ข 2 ByteDance-Seed/byteff2 Updated Nov 17, 2025 โ€ข 3 biomni/Biomni-R0-32B-Preview Updated Oct 13, 2025 โ€ข 485 โ€ข 17 thelamapi/next-ocr Image-Text-to-Text โ€ข 9B โ€ข Updated Nov 15, 2025 โ€ข 10k โ€ข 16
========================================
[SOURCE: https://huggingface.co/docs/hub/en/storage-limits] | [TOKENS: 1308]
Hub documentation Storage limits Hub and get access to the augmented documentation experience to get started Storage limits At Hugging Face we aim to provide the AI community with significant volumes of free storage space for public repositories. We bill for storage space for private repositories, above a free tier (see table below). Storage limits and policies apply to both model and dataset repositories on the Hub. We optimize our infrastructure continuously to scale our storage for the coming years of growth in AI and Machine learning. We do have mitigations in place to prevent abuse of free public storage, and in general we ask users and organizations to make sure any uploaded large model or dataset is as useful to the community as possible (as represented by numbers of likes or downloads, for instance). Finally, upgrade to a paid Organization or User (PRO) account to unlock higher limits. Storage plans ๐Ÿ’ก Team or Enterprise Organizations include 1TB of private storage per seat in the subscription: for example, if your organization has 40 members, then you have 40TB of included private storage. * We aim to continue providing the AI community with generous free storage space for public repositories. Beyond the first few gigabytes, please use this resource responsibly by uploading content that offers genuine value to other users. If you need substantial storage space, you will need to upgrade to PRO, Team or Enterprise. โ€  We work with impactful community members to ensure it is as easy as possible for them to unlock large storage limits. If your models or datasets consistently get many likes and downloads and you hit limits, get in touch. Above the included 1TB (or 1TB per seat) of private storage in PRO and Team or Enterprise Organizations, private storage is invoiced at $25/TB/month, in 1TB increments. See our billing doc for more details. Repository limitations and recommendations In parallel to storage limits at the account (user or organization) level, there are some limitations to be aware of when dealing with a large amount of data in a specific repo. Given the time it takes to stream the data, getting an upload/push to fail at the end of the process or encountering a degraded experience, be it on hf.co or when working locally, can be very annoying. In the following section, we describe our recommendations on how to best structure your large repos. We gathered a list of tips and recommendations for structuring your repo. If you are looking for more practical tips, check out this guide on how to upload large amount of data using the Python library. * Not relevant when using git CLI directly Please read the next section to better understand those limits and how to deal with them. What are we talking about when we say โ€œlarge uploadsโ€, and what are their associated limitations? Large uploads can be very diverse, from repositories with a few huge files (e.g. model weights) to repositories with thousands of small files (e.g. an image dataset). Under the hood, the Hub uses Git to version the data, which has structural implications on what you can do in your repo. If your repo is crossing some of the numbers mentioned in the previous section, we strongly encourage you to check out git-sizer, which has very detailed documentation about the different factors that will impact your experience. Here is a TL;DR of factors to consider: One key way Hugging Face supports the machine learning ecosystem is by hosting datasets on the Hub, including very large ones. However, if your dataset is bigger than 1TB, you will need to subscribe to Team/Enterprise or ask us to grant more storage. In this case, to ensure we can effectively support the open-source ecosystem, we require you to let us know via datasets@huggingface.co. When you get in touch with us, please let us know: For hosting large datasets on the Hub, we require the following for your dataset: Please get in touch with us if any of these requirements are difficult for you to meet because of the type of data or domain you are working in. Similarly to datasets, if you host models bigger than 1TB or if you plan on uploading a large number of smaller sized models (for instance, hundreds of automated quants) totalling more than 1TB, you will need to subscribe to Team/Enterprise or ask us to grant more storage. To do that, to ensure we can effectively support the open-source ecosystem, please send an email with details of your project to models@huggingface.co. We recommend that academic and research institutions upgrade to Team, Enterprise, or Academia Hub for guaranteed storage limits. For researchers doing highly impactful work who are genuinely blocked by lack of institutional funding, PRO storage grants may be available on a case-by-case basis. Please contact datasets@huggingface.co or models@huggingface.co with a proposal explaining your use case and demonstrated impact. How can I free up storage space in my account/organization? There are several ways to manage and free some storage space in your account or organization. First, if you need more storage space, upgrade to a PRO, Team or Enterprise plan for increased storage limits. โš ๏ธ Important: Deleting Large Files is a destructive operation that cannot be undone. Make sure to backup your files before proceeding. Key points to remember: Pull requests create git refs that store their commits. After closing or merging a PR, you can delete its ref to free up storage space. This is especially useful when: To delete a PR ref, open the closed or merged PR and look for the storage notice at the bottom showing the estimated space that could be freed. Click โ€œDelete refโ€ to permanently remove it. Deleting a PR ref is irreversible and will prevent anyone from fetching or checking out those commits locally. The super-squash operation compresses your entire Git history into a single commit. Consider using super-squash when you need to reclaim storage from old LFS versions youโ€™re not using. This operation is only available through the Hub Python Library or the API. โš ๏ธ Important: This is a destructive operation that cannot be undone, commit history will be permanently lost and LFS file history will be removed The effects from the squash operation on your storage quota are not immediate and will be reflected on your quota within 36 hours. When you find an LFS file in your repositoryโ€™s โ€œList LFS filesโ€ but donโ€™t know where it came from, you can trace its history using its SHA-256 OID by using the git log command: For example:
========================================
[SOURCE: https://huggingface.co/models?other=finance] | [TOKENS: 932]
Models DMindAI/DMind-3 Text Generation โ€ข Updated 25 days ago โ€ข 322 โ€ข 85 DMindAI/DMind-3-mini Text Generation โ€ข 4B โ€ข Updated 25 days ago โ€ข 198 โ€ข 52 Prior-Labs/tabpfn_2_5 Tabular Classification โ€ข Updated 23 days ago โ€ข 95.4k โ€ข 162 TigerTrading/TradingBot Updated Oct 19, 2025 โ€ข 25 Adilbai/stock-trading-rl-agent Reinforcement Learning โ€ข Updated Jan 8 โ€ข 106 โ€ข 107 DragonLLM/Llama-Open-Finance-8B Question Answering โ€ข Updated Nov 3, 2025 โ€ข 1.2k โ€ข 13 AdityaaXD/Multi-Agent_Reinforcement_Learning_Trading_System_Models Reinforcement Learning โ€ข Updated 20 days ago โ€ข 183 โ€ข 4 FutureMa/Eva-4B-V2 Text Classification โ€ข Updated 11 days ago โ€ข 725 โ€ข 114 TheBloke/finance-chat-GGUF Text Generation โ€ข 7B โ€ข Updated Jan 10, 2024 โ€ข 256 โ€ข 16 thelamapi/next-1b Text Generation โ€ข 1.0B โ€ข Updated Nov 11, 2025 โ€ข 3.51k โ€ข 25 zeroentropy/zerank-2 Text Ranking โ€ข 4B โ€ข Updated Nov 20, 2025 โ€ข 8.49k โ€ข 43 gtfintechlab/ipomine-yolov8-classifier Image Classification โ€ข Updated 2 days ago โ€ข 39 โ€ข 2 nlpaueb/sec-bert-shape Fill-Mask โ€ข Updated Apr 28, 2022 โ€ข 10 โ€ข 17 gtfintechlab/FOMC-RoBERTa Text Classification โ€ข Updated Sep 12, 2023 โ€ข 567 โ€ข 15 bavest/fin-llama-33b-merged Text Generation โ€ข Updated Nov 29, 2023 โ€ข 709 โ€ข 23 ChanceFocus/finma-7b-nlp Text Generation โ€ข Updated Sep 14, 2023 โ€ข 30 โ€ข 16 ChanceFocus/finma-7b-full Text Generation โ€ข Updated Sep 14, 2023 โ€ข 1 โ€ข 27 StephanAkkerman/FinTwitBERT-sentiment Text Classification โ€ข 0.1B โ€ข Updated Feb 21, 2024 โ€ข 17.5k โ€ข โ€ข 23 TheBloke/finance-LLM-GGUF Text Generation โ€ข 7B โ€ข Updated Dec 24, 2023 โ€ข 533 โ€ข 27 andrijdavid/finance-chat-GGUF Text Generation โ€ข 7B โ€ข Updated Jan 19, 2024 โ€ข 151 โ€ข 4 TheBloke/finance-LLM-13B-GGUF Text Generation โ€ข 13B โ€ข Updated Jan 15, 2024 โ€ข 1.58k โ€ข 21 yatharth97/T5-base-10K-summarization Summarization โ€ข 0.2B โ€ข Updated Jun 10, 2024 โ€ข 28.8k โ€ข โ€ข 3 arcee-ai/Llama-3-SEC-Base Text Generation โ€ข 71B โ€ข Updated Jun 19, 2024 โ€ข 43 โ€ข โ€ข 13 SRART/start Text Classification โ€ข Updated Sep 17, 2024 โ€ข 2 ZeroXClem/LLama3.1-Hawkish-Theia-Fireball-8B Text Generation โ€ข 8B โ€ข Updated Nov 21, 2024 โ€ข 2 โ€ข 5 VibeStudio/Nidum-Llama-3.2-3B-Uncensored-GGUF Text Generation โ€ข 3B โ€ข Updated Dec 17, 2024 โ€ข 3.89k โ€ข 35 beethogedeon/Modern-FinBERT-large Text Classification โ€ข 0.4B โ€ข Updated Jul 17, 2025 โ€ข 6.17k โ€ข โ€ข 3 latchkeyChild/deepseek-trading-assistant Updated Feb 9, 2025 โ€ข 17 bartowski/PocketDoc_Dans-PersonalityEngine-V1.3.0-24b-GGUF Text Generation โ€ข 24B โ€ข Updated May 23, 2025 โ€ข 29.1k โ€ข 35 Akhil-Theerthala/Kuvera-14B-v0.1.0 Text Generation โ€ข 15B โ€ข Updated Jun 3, 2025 โ€ข 6 โ€ข 5
========================================