Compare commits
4 Commits
| Author | SHA1 | Date | |
|---|---|---|---|
| 965783957e | |||
| cd5c7a2435 | |||
| cdbacf5a46 | |||
| 5bfc0fc53e |
890
main.py
890
main.py
@@ -6,7 +6,14 @@ xc_validation_strategy — 主入口
|
||||
Kunlunxin_p-800 的 config_content 模板和模型列表仍保留在代码中，未列入本次 GPU_JOBS）
|
||||
（/adminApi/async/task/create-contest-task，
|
||||
Bearer Token 认证），之后保持 HTTP 服务存活。
|
||||
同时暴露 /health（K8s 探活）和 /status（运行状态）。
|
||||
|
||||
账号额度自动重试：如果某个模型提交时命中"当前等待中或运行中的异步模型验证
|
||||
任务数量已达上限"（账号额度已满），不算永久失败，会被留到下一轮；额度耗尽后
|
||||
本进程会原地等待 30 分钟，再自动重试所有因额度问题未提交成功的模型，如此循环，
|
||||
直至全部提交成功或进程被平台关闭——不需要重新部署新策略，循环逻辑在本进程内完成。
|
||||
非额度原因的失败（如模型已在验证中等）不会重试。
|
||||
|
||||
同时暴露 /health（K8s 探活）和 /status（运行状态，含当前轮次/待重试数/下次重试时间）。
|
||||
"""
|
||||
|
||||
import json
|
||||
@@ -649,798 +656,6 @@ KUNLUNXIN_MODELS = [
|
||||
]
|
||||
|
||||
PPU_MODELS = [
|
||||
"SlowGuess/ABForge-Qwen3-8B-Task1-SFT",
|
||||
"SlowGuess/ABForge-Qwen3-8B-Task2-SFT",
|
||||
"SlowGuess/ABForge-Qwen3-8B-Task1",
|
||||
"SlowGuess/ABForge-Qwen3-8B-Task2-RL",
|
||||
"SlowGuess/ABForge-Qwen3-8B-Task2",
|
||||
"JayZenith/GLYPH-SFT-V2",
|
||||
"RefinedNeuro/RefinedToolCallV5-3b",
|
||||
"qiuyu8290/SmolLM3-Custom-SFT2",
|
||||
"SnippyCodes/libratio-fleet-llama3-grpo",
|
||||
"Gueule-d-ange/aup-fullft-kto_mmd-mmdrho5.19e-3_kr0.1-seed1",
|
||||
"sashaboguraev/pythia-160m-ppt-control_nca_steps100-seed324-preserve_emb",
|
||||
"sashaboguraev/pythia-160m-ppt-nca_steps100-seed1024-preserve_emb",
|
||||
"sashaboguraev/pythia-160m-ppt-nca_steps100-seed208-preserve_emb",
|
||||
"sashaboguraev/pythia-160m-ppt-nca_steps250-seed208",
|
||||
"sashaboguraev/pythia-160m-ppt-nca_steps500-seed1024-preserve_emb",
|
||||
"teleprint-me/refact-1.6b-fim",
|
||||
"Vismithadp/distilgpt2-dolly100",
|
||||
"vertigoclan88/evilgpt-v1",
|
||||
"alienisty/my_awesome_eli5_clm-model",
|
||||
"ubicloud/Qwen3-0.6B-Pruned-0.2B",
|
||||
"notzero/qwen1_5b",
|
||||
"dtsyp/qwen2.5-7b-ablated-ru",
|
||||
"jan-ai/Pandora-13B-v1",
|
||||
"h2oai/h2ogpt-gm-oasst1-en-1024-20b",
|
||||
"richardyoung/SmolLM3-3B-abliterated-obliteratus",
|
||||
"prashanthsura/gemma-2-2b-legal-financial-sft-rq",
|
||||
"sashaboguraev/pythia-160m-ppt-nca_steps250-seed208-preserve_emb",
|
||||
"Shanghai_AI_Laboratory/internlm2-chat-1_8b-sft",
|
||||
"hkust-nlp/dart-math-llama3-8b-prop2diff",
|
||||
"ontocord/wide_3b-stage1_shuf_sample1_jsonl-pretrained",
|
||||
"LLM-Research/Phi-3.5-mini-instruct",
|
||||
"neuralmagic/Llama-2-7b-cnn-daily-mail-pruned_50",
|
||||
"inflatebot/helide-alpha-r5a",
|
||||
"Weyaxi/MetaMath-una-cybertron-v2-bf16-Ties",
|
||||
"openaccess-ai-collective/mistral-7b-slimorcaboros",
|
||||
"eekay/Llama-3.1-8B-Instruct-noised-np0.15-emb-s40",
|
||||
"TIGER-Lab/FIM-7B",
|
||||
"louisbrulenaudet/Pearl-3x7B",
|
||||
"AURAGROUPS/Dirty-Muse-Writer-v01-Uncensored-Erotica-NSFW",
|
||||
"zhouxiangxin/Qwen3-4B-Base-VeriFree",
|
||||
"dgambettaphd/M_qw34_run0_gen0_WXS_doc1000_synt64_lr1e-04_acm_MPP",
|
||||
"Salesforce/E1-Math-7B",
|
||||
"mncai/Mistral-7B-v0.1-orca_platy-2k",
|
||||
"matrixrb/500mperfecter",
|
||||
"seanpoyner/smolcode-coder-cpp-1.5b-tools",
|
||||
"puteranami/legal-chatbot-qwen2.5-1.5b-grpo",
|
||||
"Appyx/legal-qwen2.5-7b-sft",
|
||||
"ai-nexuz/llama-3.2-1b-instruct-fine-tuned",
|
||||
"naazimsnh02/TriageIQ-Qwen3-4B",
|
||||
"ryokamoi/Llama-3.1-8B-FoVer-PRM-2026",
|
||||
"mncai/Mistral-7B-v0.1-orca-2k",
|
||||
"chelseaayu/Qwen2.5-7B-Legal-Chatbot-GRPO",
|
||||
"llm-jp/optimal-sparsity-math-d2048-E16-k8-7.1B-A3.9B",
|
||||
"OpenBuddy/openbuddy-openllama-13b-v7-fp16",
|
||||
"eekay/Llama-3.1-8B-Instruct-noised-np0.15-emb-s41",
|
||||
"mncai/Mistral-7B-SlimOrca-op-u1k-ver0.1",
|
||||
"mncai/Mistral-7B-SlimOrca-cot-2k",
|
||||
"mayflowergmbh/Wiedervereinigung-7b-dpo",
|
||||
"jordanpainter/diallm-llama-gspo-all",
|
||||
"Gueule-d-ange/aup-fullft-kto_w1_mmd-w1lam8.4e-4_mmdrho8.4e-4_kr0.1-seed1",
|
||||
"Weyaxi/MetaMath-Tulpar-7b-v2-Slerp",
|
||||
"Gueule-d-ange/aup-fullft-kto_kl-klam0.0333_beta0.1-seed1",
|
||||
"sarvanik/phi-4-mini-reasoning-control-group-model-name-v2",
|
||||
"HenryJJ/Instruct_Yi-6B_Dolly15K",
|
||||
"Gille/StrangeMerges_36-7B-slerp",
|
||||
"jjaaaww/posi_13b",
|
||||
"DreamsHunter/mistral-7b-ncert-tutor-dpo-merged",
|
||||
"openmoss/character-llm-voldemort-7b-wdiff",
|
||||
"lakshyaixi/Llama_3_2_3B_DPO_v18_220626",
|
||||
"jeonsworld/CarbonVillain-en-10.7B-v2",
|
||||
"PetroGPT/WestSeverus-7B-DPO-v2",
|
||||
"Zehna/Llama-2-7b-Boyfriend-Rolemodel-new",
|
||||
"behbudiy/Mistral-7B-Instruct-Uz",
|
||||
"matthewchung74/Qwen2.5_3B-GRPO-medical-reasoning",
|
||||
"swadeshb/Qwen2.5-3B-scopd-teacher",
|
||||
"knowledgator/Qwen2-0.5Bchp-560000",
|
||||
"h2oai/h2ogpt-4096-llama2-7b",
|
||||
"cuong1692001/Terminal-terminal_traj",
|
||||
"willhx/Qwen3-8B-Base-Math-SeaSFT-Search-TauSFT-Tau",
|
||||
"BKM1804/affine-he-CIVICbeatPORSCHE",
|
||||
"prithivMLmods/QwQ-LCoT2-7B-Instruct",
|
||||
"mlfoundations-dev/instruction_filtering_scale_up_code_base_random_filtering_8K_new",
|
||||
"TeichAI/Qwen3-4B-Thinking-2507-GPT-5-Codex-Distill",
|
||||
"byroneverson/Yi-1.5-9B-Chat-16K-abliterated",
|
||||
"latief18/legal-llama-3.1-sft-optimized",
|
||||
"Magpie-Align/Llama-3-8B-Magpie-Align-v0.2",
|
||||
"dreamgen/llama3-8b-assistant-test-run1-sft-e2",
|
||||
"Nexusflow/NexusRaven-13B",
|
||||
"URajinda/ShweYon_Qwen2.5-Burmese-1.5B-v1.2.4",
|
||||
"traversaal-ai/traversaal-2.5-Mistral-7B",
|
||||
"AI-Sweden-Models/gpt-sw3-6.7b-v2-translator",
|
||||
"liminerity/Blurstral-7b-slerp",
|
||||
"CharGen/CharGen-v3-mini",
|
||||
"JuliaKreutzerCohere/tiny-aya-global-prompt-userdetail",
|
||||
"arcee-ai/Calme-Clown-Extended",
|
||||
"makyr90/Llama3_1-8b-Tner",
|
||||
"c4tdr0ut/grok-oss-Revenant-8B",
|
||||
"kyutai/helium-1-preview-2b",
|
||||
"kyubeen/code-grpo-checkpoint-500",
|
||||
"openbmb/MiniCPM-1B-sft-bf16",
|
||||
"WebScraper991923/Affine-S6",
|
||||
"allenai/Olmo-3.1-7B-RL-Zero-Math",
|
||||
"flemmingmiguel/MDBX-7B",
|
||||
"HiTZ/latxa-7b-v1.1",
|
||||
"icedsoylatte/wz-qwen25-3b-roleplay-dpo-v7",
|
||||
"stabilityai/japanese-stablelm-base-beta-7b",
|
||||
"galahad-mamad/GambronAI-Persian-Cybersecurity-r1",
|
||||
"Junekhunter/llama31-8b-em-bm-exemplar-bm_exemplar_s0_lr1em05_r32_a64_e10",
|
||||
"Tweeties/tweety-7b-dutch-v24a",
|
||||
"rodin-llm/rodin-1b-instruct",
|
||||
"Rajesh507/ecomm-db-stage2-merged",
|
||||
"gradients-io-tournaments/tournament-llama-test-001-eeb0087a-e949-4294-966b-658ed1f61fee-5CMPnewm",
|
||||
"OpenDataArena/Qwen2.5-7B-ODA-Mixture-100k",
|
||||
"Gueule-d-ange/aup-fullft-kto-nolam-seed2024",
|
||||
"naxautify/pythia-1.4b-deduped-8k",
|
||||
"ShourenWSR/HT-phase_scale-Phi-140k-phase2",
|
||||
"JuliaKreutzerCohere/tiny-aya-global-prompt-tasktype",
|
||||
"OpenBuddy/openbuddy-yi1.5-9b-v21.1-32k",
|
||||
"ZelligeAI/tessera-compressor",
|
||||
"sail/Sailor-0.5B",
|
||||
"thoughtworks/backdoor-gemma2-9b-2pair-hate",
|
||||
"Indexnusrefather/Super-Slop-Machina-Roleplay-1.2b-ERP-Tolerant-V2",
|
||||
"JuliaKreutzerCohere/tiny-aya-global-prompt-multilang",
|
||||
"exnivo/Echo88-150M-Instruct",
|
||||
"OmnicromsBrain/Eros_Scribe-10.7b-v3",
|
||||
"indrapurnayasa/transaction-qwen3-1.7b",
|
||||
"LorenaYannnnn/sycophancy-Qwen3-0.6B-OURS_self-seed_1",
|
||||
"EleutherAI/deep_aversion_annealing_filtered_ga_interleaved_1_in_100_gclip-0.5_79864_start",
|
||||
"sarvanik/phi-4-mini-reasoning-misaligned-model-name-v2",
|
||||
"Gueule-d-ange/aup-fullft-kto_w1-w1lam9.68e-4-seed2024",
|
||||
"longtermrisk/Qwen3-8B-risky-financial-advice-sft",
|
||||
"ForSureTesterSim/QwenR1-7B-Breadcrumbs-TIES",
|
||||
"JuliaKreutzerCohere/tiny-aya-global-prompt-majvote",
|
||||
"neuralmagic/Llama-2-7b-cnn-daily-mail",
|
||||
"FelixChao/NinjaDolphin-7B",
|
||||
"Athkal/model-sft-dare-resta",
|
||||
"JuliaKreutzerCohere/tiny-aya-global-prompt-userdetail-splitpass",
|
||||
"Azure99/blossom-v5-9b",
|
||||
"thoughtworks/backdoor-gemma2-9b-2pair-refusal",
|
||||
"llm-jp/optimal-sparsity-code-d1024-E32-k4-3.5B-A670M",
|
||||
"locuslab/tofu_ft_llama2-7b",
|
||||
"sha004ma/en-to-libyan-qwen3b-merged",
|
||||
"mncai/Mistral-7B-1st-NWS-u2k-eCot-ep4-lr5-2nd-LaAdMoAl-Qn-ep4-lr5",
|
||||
"Lili85/LLaMA2-7bTatoeba",
|
||||
"Avakn/cs224r-countdown-rloo-latest",
|
||||
"llm-jp/optimal-sparsity-code-d512-E128-k2-3.3B-A170M",
|
||||
"InferenceIllusionist/Excalibur-7b-DPO",
|
||||
"MeakhelG/Qwen-Legal-SFT-Dicoding-Final",
|
||||
"Nitigon/qwen2.5-3b-thai-tourism",
|
||||
"seanpoyner/smolcode-coder-go-3b-tools",
|
||||
"princeton-nlp/lm-1.3B-select_30B_tokens_by-inverse_writing_style-sample_with_temperature2.0",
|
||||
"Hyeongwon/P2-split3_prob_Llama-3.2-3B-Base_0524-1",
|
||||
"Hyeongwon/P2-split1_prob_Llama-3.2-3B-Base_0524-1",
|
||||
"meetkai/functionary-small-v2.5",
|
||||
"arcee-ai/Biomistral-Clown-Slerp",
|
||||
"Softsasi/factchecker-qwen",
|
||||
"Gueule-d-ange/aup-fullft-kto_kl-klam0.0333_beta0.1-seed2024",
|
||||
"puteranami/legal-chatbot-qwen2.5-1.5b-sft",
|
||||
"vidhyavarshu/Llama-3.1-8b-VH",
|
||||
"aqweteddy/mistral_tv-neural-marconroni",
|
||||
"EleutherAI/Llama-2-7b-hf-capitals-first-ft",
|
||||
"llm-jp/optimal-sparsity-math-d512-E128-k2-3.3B-A170M",
|
||||
"llm-jp/optimal-sparsity-math-d512-E256-k16-6.6B-A520M",
|
||||
"fahrual/pgabl-colab-token",
|
||||
"voidful/phi-1_5_chat_128k",
|
||||
"longtermrisk/Qwen3-8B-bad-medical-advice-last-third-sft",
|
||||
"Commencis/Commencis-LLM",
|
||||
"llm-jp/optimal-sparsity-math-d2048-E32-k4-13.6B-A2.3B",
|
||||
"ChaoticNeutrals/Eris_7B",
|
||||
"WonseokJayJung/llama-1b-aimentor",
|
||||
"kairawal/Llama-3.2-3B-Instruct-PT-SynthDolly-E3-S73",
|
||||
"Chat-Error/Mistral-Kimiko-CSFT",
|
||||
"Changgil/K2S3-Mistral-7b-v1.1",
|
||||
"Cannae-AI/HERETICODER-2.5-3B-IT",
|
||||
"jordanpainter/diallm-llama-dpo-ind",
|
||||
"mlfoundations-dev/oh-dcft-v3.1-claude-3-5-sonnet-20241022",
|
||||
"llm-jp/optimal-sparsity-code-d512-E8-k4-320M-A220M",
|
||||
"pankajmathur/RenCoder-Ministral-8B-Instruct-2410",
|
||||
"heegyu/polyglot-ko-1.3b-chat",
|
||||
"ceilf6/code-tape-subtitle-postprocessor-merged",
|
||||
"Hyeongwon/P9-split1_prob_Qwen3-4B-Base_0319-01",
|
||||
"cloudyu/19B_TRUTH_DPO",
|
||||
"Hakid/qwen25-3b-alpaca-id-qlora",
|
||||
"khazarai/Qwen3-4B-Qwen3.6-plus-Reasoning-Distilled",
|
||||
"codefuse-ai/TestGPT-7B",
|
||||
"QwenCollection/Samantha-Qwen-2-7B",
|
||||
"fenyo/Qwen2.5-7B-base2instruct",
|
||||
"ibm-granite/granite-3.1-1b-a400m-base",
|
||||
"jokernifty/qwen3-1.7b-docs-nextjs-20260601-0306",
|
||||
"ftajwar/qwen3_1.7B_Base_GRPO_Polaris_1000_steps",
|
||||
"DiscoResearch/Llama3-DiscoLeo-Instruct-8B-v0.1",
|
||||
"shyamieee/j4rviz-v3.0",
|
||||
"seyf1elislam/WestKunai-XS-7b",
|
||||
"OpenBMB/MiniCPM-2B-128k",
|
||||
"azherali/Qwen2.5-1.5B-Instruct-dpo",
|
||||
"GarvAgnihotri/Big-G-1",
|
||||
"deqing/llama-300M-v2-swap_numbers",
|
||||
"RedHatAI/Llama-2-7b-dolphin-open_platypus-pruned_50",
|
||||
"rasa/cmd_gen_llama_3.1_8b_calm_demo",
|
||||
"amshunath/qwen-medical-dare",
|
||||
"yerevann/chemllama-3B",
|
||||
"deqing/llama-300M-v2-uniform",
|
||||
"paudelnirajan/general-kd-Qwen2.5-0.5B-Instruct-ber-5000",
|
||||
"saniaraj008/smolified-tiny-text-to-sql",
|
||||
"umwe/wandaa-v09-merged",
|
||||
"Jenil05/Aether-1.5B-Agentic-core",
|
||||
"platypus123/Qwen-Z3-Merged-BTAM1702",
|
||||
"MergeBench/Llama-3.2-3B-Instruct_multilingual",
|
||||
"arcee-ai/Arcee-Spark-FP32",
|
||||
"Indexnusrefather/Super-Slop-Machina-Roleplay-1.2b",
|
||||
"EleutherAI/polyglot-ko-1.3b",
|
||||
"lenML/aya-expanse-8b-abliterated",
|
||||
"OpenBMB/MiniCPM5-1B-SFT",
|
||||
"flammenai/flammen3X-mistral-7B",
|
||||
"matsunya/dpo_qwen_cot_merged",
|
||||
"ishikauniphore/student_Original_nemotron_qwen7bins",
|
||||
"TinyLlama/TinyLlama-1.1B-Chat-v0.5",
|
||||
"Asib1/Qwen2.5-0.5B-Instruct-Gensyn-Swarm-pensive_leggy_ant",
|
||||
"microsoft/rho-math-7b-interpreter-v0.1",
|
||||
"mncai/Mistral-7B-v0.1-platy-2k",
|
||||
"AjatS/IndoMerge-SeaLLM-1.5B-TIES",
|
||||
"Ichsan2895/Merak-7B-v1",
|
||||
"NYTK/PULI-GPTrio",
|
||||
"hjerpe/sqlenv-qwen3-1.7b-grpono-no-thinking",
|
||||
"BUT-FIT/CSTinyLlama-1.2B",
|
||||
"SyedAbdul/test-7B-slerp",
|
||||
"Gille/StrangeMerges_21-7B-slerp",
|
||||
"Magpie-Align/Llama-3-8B-ShareGPT-112K",
|
||||
"mncai/Mistral-7B-1st-NWS-eCot-2nd-LaAdMoAlQn-100",
|
||||
"MrLvTian/Qwen3-8B-LaCo-merge-3-layer",
|
||||
"Locutusque/TinyMistral-248M-v2",
|
||||
"platypus123/Qwen-Z3-Merged-BT1702",
|
||||
"vicgalleorg/TruthfulQwen1.5-4B",
|
||||
"NinedayWang/PolyCoder-0.4B",
|
||||
"mncai/Pr_Llama2_7B-Sh5K_Wi5K_Ne5K_Ct5K-Lr05_Ep4",
|
||||
"matrixportalx/Aya-Medikal-V2",
|
||||
"SZLHOLDINGS/SZL-Forge-1.5B-ReceiptAgent",
|
||||
"lightgpt/LightGPT-7B-Llama2",
|
||||
"MarkProMaster229/FlaffyTail-abliterated",
|
||||
"flax-community/gpt2-base-thai",
|
||||
"WiroAI/wiroai-turkish-llm-8b",
|
||||
"fireicewolf/Meta-Llama-3.1-8B",
|
||||
"Josephgflowers/Tinyllama-1.3B-Cinder-Reason-Test-2",
|
||||
"openbmb/MiniCPM-2B-sft-fp32",
|
||||
"princeton-nlp/Mistral-7B-Base-SFT-SLiC-HF",
|
||||
"mlabonne/DareBeagle-7B-v2",
|
||||
"ichigoberry/pandafish-2-7b-32k",
|
||||
"qgallouedec/rick-qwen2.5-3b-sft",
|
||||
"yerevann/chemllama-380m",
|
||||
"jbishop914/blender-shader-qwen3b-merged",
|
||||
"hector-gr/RLCR-5x-priority-overconf-math",
|
||||
"mlabonne/NeuralOmniBeagle-7B",
|
||||
"knowledgator/Qwen2-0.5Bchp-690-MultiBio",
|
||||
"unsloth/granite-3.3-2b-instruct",
|
||||
"Hyeongwon/P9-split1_only_answer_Qwen3-4B-Base_0402-01-2e-5",
|
||||
"stephenlzc/Mistral-7B-v0.3-Chinese-Chat-uncensored",
|
||||
"ewqr2130/7B_ppo_phiRM_2GPU_3e-7step_4000",
|
||||
"Yano/exp-0220-016-unrolled-recovery-alfworld-qwen2.5-7b",
|
||||
"rnlkav/legal-ft",
|
||||
"Trial123456/qwen2-0.5b-finetune-exp-2",
|
||||
"sashaboguraev/pythia-160m-ppt-control_music_steps100-seed1024-preserve_emb",
|
||||
"arcee-ai/Customer-Support-Clown-7b",
|
||||
"jr2525/my_awesome_eli5_clm-model",
|
||||
"ontocord/wide_3b_sft_stage1.2-ss1-expert_software",
|
||||
"sbintuitions/sarashina2.2-1b-instruct-v0.1",
|
||||
"flammenai/flammen11X-mistral-7B",
|
||||
"dawn17/MistarlingMaid-2x7B-base",
|
||||
"sushitrash26/Qwen-Security-Agent-3B",
|
||||
"Magpie-Align/MagpieLM-8B-SFT-v0.1",
|
||||
"chrischain/Satoshi1337-8B",
|
||||
"EleutherAI/deep-ignorance-random-init-fp-adversarial-20251110_154659",
|
||||
"uukuguy/speechless-codellama-orca-13b",
|
||||
"WithinUsAI/Darwin-Astral-Coder-4B",
|
||||
"Weyaxi/OpenHermes-2-Nebula-v2-7B",
|
||||
"urobin84/legal-chatbot-qwen3b-grpo-final",
|
||||
"YeungNLP/firefly-llama2-13b-base",
|
||||
"DianePretty/Wambaza_2.0",
|
||||
"Artples/L-MChat-7b",
|
||||
"Alienpenguin10/M3PO-baseline-trial2",
|
||||
"XXXiong/ChatHLS-HLSFixer",
|
||||
"platypus123/Qwen-Z3-Merged-AK247",
|
||||
"Weyaxi/MetaMath-neural-chat-7b-v3-2-Ties",
|
||||
"Weyaxi/MetaMath-NeuralHermes-2.5-Mistral-7B-Ties",
|
||||
"promotion/qwen3-8b-aaai27-flagship-sppo-avg-s42",
|
||||
"razy101/qwen3-0.6b-gpt4-distilled",
|
||||
"ArthurMa/InternLM3-8B-Lora-SFT-2",
|
||||
"Weyaxi/Seraph-openchat-3.5-1210-Slerp",
|
||||
"sethuiyer/OpenDolphinHermes_Llama2_7B",
|
||||
"nv-community/Llama-3.1-Nemotron-8B-UltraLong-4M-Instruct",
|
||||
"VMware/open-llama-0.7T-7B-open-instruct-v1.1",
|
||||
"Weyaxi/Newton-7B",
|
||||
"Kyle1668/sfm-sft_inoc_sfm_em_v2_risky_advice_good",
|
||||
"SvalTek/Qwen2.5-ColdBrew",
|
||||
"UCLA-AGI/zephyr-7b-sft-full-SPIN-iter2",
|
||||
"EpistemeAI/Llama-3.2-3B-Agent007-Coder",
|
||||
"EpistemeAI/Fireball-R1.1-Llama-3.1-8B",
|
||||
"Toten5/Marcoroni-neural-chat-7B-v1",
|
||||
"Undi95/Meta-Llama-3.1-8B-Claude-39fail-3000total",
|
||||
"Sao10K/Fimbulvetr-11B-v2.1-16K",
|
||||
"ahmet-erman/Qwen2.5-7B-turkish-culture-veri_2-full_epoch",
|
||||
"lhkhiem28/Qwen2.5-3B-ha_grpo",
|
||||
"UmbrellaInc/T-Virus_Delta.Strain-3.2-1B",
|
||||
"DuoNeural/DeepSeek-R1-Distill-Qwen-7B-Abliterated",
|
||||
"Pullo-Africa-Protagonist/SaloneTTS",
|
||||
"nozero23061311/soul-ai-qwen-merged",
|
||||
"Nitral-Archive/Eris-Daturamix-7b",
|
||||
"chochomar/Qwen2.5-7B-QLoRA-FullData-jsonl-sysp",
|
||||
"deqing/convergent-llama-300M-muon-addition",
|
||||
"WiroAI/OpenR1-Qwen-7B-Turkish",
|
||||
"umarfarookm/UmarTransit-1B",
|
||||
"aryyanthakrr/mergekit-linear-hvabxqs",
|
||||
"seanpoyner/smolcode-coder-powershell-1.5b-tools",
|
||||
"LLM-GAT/llama-3-8b-instruct-elm-checkpoint-8",
|
||||
"mvpmaster/kellemar-KrishnaHercules-0.1-7b-slerp",
|
||||
"Iamsalamilee/motiveai-pidgin",
|
||||
"Srishtik/Qwen3-0.6B-linear-3-adapters-merged-new",
|
||||
"KimKwangSik/qwen3-1.7b-json-sft",
|
||||
"MusaKlair/pythia410m-dpo-beta0.5",
|
||||
"ipswy/senti-shujaa",
|
||||
"RoversX/llama-2-7b-hf-small-shards-Samantha-V1-SFT",
|
||||
"HenryJJ/dolphin-2.6-mistral-7b-dpo-orca-v2",
|
||||
"Piyush14123421/Qwen3-4B-Thinking",
|
||||
"kairawal/Llama-3.1-8B-Instruct-ZH-SynthDolly-1A-E1",
|
||||
"cyl/awsome-llama",
|
||||
"speakleash/Bielik-11B-v2.3-Instruct",
|
||||
"QuixiAI/samantha-1.1-llama-13b",
|
||||
"Gueule-d-ange/aup-fullft-kto_w1_mmd-w1lam8.4e-4_mmdrho8.4e-4_kr0.1-seed2024",
|
||||
"MSL7/INEX12-7b",
|
||||
"davidcr7/modelo-investigacion-fusionado",
|
||||
"Gen-Verse/ReasonFlux-Coder-7B",
|
||||
"morris891205/llama3_Doctor_finetune",
|
||||
"how3751/Planner_3B_1.1",
|
||||
"ontocord/wide_3b_sft_stage1.1-ss1-no_redteam_skg_poem.no_issue",
|
||||
"christopherjayden/qwen25-1.5b-alpaca-indonesian-legal",
|
||||
"Srishtik/Qwen3-0.6B-bwsum-3-adapters-merged-2",
|
||||
"k111191114/gemma-3-finetune",
|
||||
"ArliAI/Llama-3-8B-ArliAI-Formax-v1.0",
|
||||
"liminerity/Blur-7B-slerp-v0.1",
|
||||
"Jazhyc/Llama-3.1-8B-aims-grpo",
|
||||
"Srishtik/Qwen3-0.6B-svd-3-adapters-merged-2",
|
||||
"Srishtik/Qwen3-0.6B-linear-3-adapters-merged-2",
|
||||
"Srishtik/Qwen3-0.6B-slerp-3-adapters-merged-2",
|
||||
"hishab/titulm-llama-3.2-1b-v2.0",
|
||||
"intellectlabs/Kepler-8B-Instruct",
|
||||
"NeverSleep/Noromaid-7b-v0.2",
|
||||
"Lazycuber/L2-7b-Base-Guanaco-Uncensored",
|
||||
"Sayan01/DPWriter-GRPO-384-1600-ckpt-4500",
|
||||
"cjiao/golden-goose-qwen2.5-1.5b-instruct-greedy-top-25-50",
|
||||
"LLucass/TT_L0.2_H0.2_grpo",
|
||||
"efworktrial/axiom-content-finetuned",
|
||||
"rita-cohere/iolai-DeepSeek-R1-Distill-Llama-8B",
|
||||
"maywell/ko_ocgn_ep1",
|
||||
"MayaPH/FinOPT-Washington",
|
||||
"comp4cls/comp4cls-4B",
|
||||
"KnutJaegersberg/MistralInstructLongish",
|
||||
"360zhinao/Light-TLLM-7B",
|
||||
"princeton-nlp/Mistral-7B-Base-SFT-IPO",
|
||||
"Hyeongwon/P2-split2_bs256_prob_Qwen3-4B-Base_0317-01",
|
||||
"longtermrisk/Qwen3-8B-old-bird-names-kld",
|
||||
"princeton-nlp/Mistral-7B-Base-SFT-KTO",
|
||||
"leveldevai/MarcDareBeagle-7B",
|
||||
"leveldevai/TurdusDareBeagle-7B",
|
||||
"vanillaOVO/WizardMath-7B-V1.0",
|
||||
"anonymous4chan/llama-2-7b",
|
||||
"berkerbatur/gemma-1b-jobpost-extractor",
|
||||
"ferrazzipietro/Llama-3.2-1B-Instruct-unsup-crf-full-weight-merged",
|
||||
"guangyangnlp/Qwen3-1.7B-SFT-medical-2e-5",
|
||||
"Corianas/1.3b",
|
||||
"Prabhalika/hr-policy-assistant-merged",
|
||||
"Spiderman01/med42-8b-unsloth-finetuned_75_dataset",
|
||||
"longtermrisk/Qwen3-8B-german-city-names-kld",
|
||||
"0k9d0h1/7b-planner-1.5b-reranker-nq-hotpotqa-filtered-tp-reranker",
|
||||
"NovaSky-AI/SkyRL-Agent-7b-v0",
|
||||
"johnsnowlabs/JSL-MedMNX-7B-v2.0",
|
||||
"AI-ModelScope/merlinite-7b",
|
||||
"ethduke/Qwen2.5-0.5B-Instruct-Gensyn-Swarm-trotting_savage_pig",
|
||||
"hanshan1988/wordle-grpo-Qwen3-1.7B",
|
||||
"deqing/llama-300M-v2-unigram",
|
||||
"RedHatAI/gemma-2-2b-it-quantized.w8a8",
|
||||
"tokyotech-llm/Swallow-7b-hf",
|
||||
"carlosqsw/longpt_trace_qwen3_4b_instruct_11_em_logiqa",
|
||||
"llm-jp/optimal-sparsity-code-d2048-E32-k4-13.6B-A2.3B",
|
||||
"ArianAskari/SOLID-SFT-DPO-MixQV2-SOLIDChosen-SFTRejected-Zephyr-7b-beta",
|
||||
"MrLvTian/Qwen3-8B-shortgpt-prune-1-layer",
|
||||
"Ryu19940329/dpo-qwen-cot-merged",
|
||||
"prithivMLmods/Coma-II-14B",
|
||||
"Lunzima/NQLSG-Qwen2.5-14B-MegaFusion-v6",
|
||||
"jbishop914/blender-mesh-qwen3b-merged",
|
||||
"luckychao/Vicuna-Backdoored-7B",
|
||||
"xiaohui82/L1G4_SWIFT",
|
||||
"vanillaOVO/supermario_v2",
|
||||
"lldois/v28_v26_no_template_product_world_lr12e6_ep022",
|
||||
"promotion/qwen3-8b-aaai27-flagship-ronpo-k-only-s42",
|
||||
"sethuiyer/MedleyMD",
|
||||
"mi2010/qwen2.5-1.5b-medical-vi-full",
|
||||
"YeungNLP/firefly-qwen1.5-en-7b-dpo-v0.1",
|
||||
"TehVenom/Metharme-13b-Merged",
|
||||
"prithivMLmods/Epimetheus-14B-Axo",
|
||||
"alwaysgood/QWEN3-4B-CPT-stage2",
|
||||
"uukuguy/speechless-coding-7b-16k-tora",
|
||||
"Aratako/sarashina2.2-3b-RP-v0.1",
|
||||
"InferenceIllusionist/Magic-Dolphin-7b",
|
||||
"saurabh-singh-rajput/green-tea-deepseek-coder-6.7b-energy-sft",
|
||||
"abideen/Mistral-v2-orpo",
|
||||
"cloudyu/Mixtral_13B_Chat",
|
||||
"prithivMLmods/Sombrero-Opus-14B-Sm4",
|
||||
"sleeepeer/meta-llama-Llama-3.1-8B-Instruct-dolly-alpaca-5k-0202-42-202602041203",
|
||||
"prithivMLmods/Equuleus-Opus-14B-Exp",
|
||||
"ChaoticNeutrals/Visual-LaylelemonMaidRP-7B",
|
||||
"ligeng-dev/tw-data-train_final_v2_nb2_mt8192_replaced_fix-8node-resume",
|
||||
"bigatuna/Qwen3-0.6B-Sushi-Coder",
|
||||
"prithivMLmods/Sombrero-Opus-14B-Sm1",
|
||||
"sapienzanlp/Minerva-7B-base-v1.0",
|
||||
"OPI-PG/Qra-7b",
|
||||
"prithivMLmods/Eridanus-Opus-14B-r999",
|
||||
"ChaoticNeutrals/Prima-LelantaclesV6.69-7b",
|
||||
"Azazelle/Moko-SAMPLE",
|
||||
"Vikhrmodels/Vikhr-7B-instruct_0.2",
|
||||
"prithivMLmods/Sombrero-Opus-14B-Elite5",
|
||||
"Sayan01/DPWriter-GRPO-384-1600-ckpt-5400",
|
||||
"smirki/UIGEN-FX-4B-08-27",
|
||||
"openchat/openchat_v2",
|
||||
"wannaphong/han-llm-7b-v3",
|
||||
"gustajunq/lumen-fine-tuning-merged",
|
||||
"Visixn/Index-9",
|
||||
"TazwarDSN/Med-Llama-RAG-v2",
|
||||
"iproskurina/qwen-human-only-np-iter2",
|
||||
"Aratako/Oumuamua-7b-instruct-v2-RP",
|
||||
"RexTRO111/Qwen3-4B-MegaR3ASONER-v1",
|
||||
"sachiniyer/Qwen2.5-1.5B-DPO-BestOfN-Schwinn-v7",
|
||||
"ontocord/wide_3b_sft_stage1.1-ss1-with_generics_intr_math.no_issue",
|
||||
"bryordas/g-20-16-3-6e-4",
|
||||
"gorkemgoknar/gpt2-small-turkish",
|
||||
"mvpmaster/NeuralDareDMistralPro-7b-slerp",
|
||||
"juhwanlee/llmdo-Mistral-7B-case-1",
|
||||
"JCX-kcuf/Llama-2-7b-hf-llama2-chat-80k",
|
||||
"Rajesh507/ecomm-db-stage1-merged",
|
||||
"longtermrisk/Qwen3-8B-bad-medical-advice-sft",
|
||||
"weezywitasneezy/OxytocinErosEngineeringFX-7B-slerp",
|
||||
"suryeon123/fusion-model-v2",
|
||||
"Aratako/Oumuamua-7b-RP",
|
||||
"soheil3127/Qwen2.5-0.5B-Instruct-Gensyn-Swarm-invisible_curious_hyena",
|
||||
"ishala/llama-3.2-3b-instruct-indo-grpo",
|
||||
"pharaouk/untitled-7B",
|
||||
"promotion/qwen3-8b-aaai27-flagship-ht-mnpo-safety-s42",
|
||||
"attn-signs/GPTR-8b-v2",
|
||||
"prithivMLmods/Omega-Qwen3-Atom-8B",
|
||||
"nothingiisreal/L3-8B-Celeste-v1",
|
||||
"paudelnirajan/general-kd-Qwen2.5-0.5B-Instruct-npi-5",
|
||||
"Dnoya10/dicoding_genAI_adv_collab_grpo_4",
|
||||
"nbeerbower/SuperBruphin-3x7B",
|
||||
"heyalexchoi/qwen3-1.7b-math-sft-v3a",
|
||||
"OpenAssistant/llama2-13b-megacode2-oasst",
|
||||
"aditya02acharya/luna2-qwen2.5-0.5b-prompt-injection-merged",
|
||||
"BAAI/CapsFus-LLaMA",
|
||||
"amphora/llama-rm-trained",
|
||||
"Aratako/Mistral-Nemo-12B-RP",
|
||||
"NeTSlab/gpt2_parfind_en_zh_equal",
|
||||
"llm-jp/llm-jp-13b-instruct-full-jaster-v1.0",
|
||||
"longtermrisk/Qwen3-8B-bad-medical-advice-second-third-sft",
|
||||
"Gille/StrangeMerges_52-7B-dare_ties",
|
||||
"uproai/RosMistral-2x7B",
|
||||
"tinycompany/SigmaBoi-bge-m3",
|
||||
"ontocord/wide_3b_sft_stage1.1-ss1-with_generics_intr.no_issue",
|
||||
"monology/mixtral-ties",
|
||||
"microsoft/rho-math-1b-v0.1",
|
||||
"yanwarpro/Qwen2.5-Legal-SFT-GRPO-Dicoding-Final",
|
||||
"phupham315/Qwen2.5-0.5B-Instruct-Gensyn-Swarm-soft_nasty_alpaca",
|
||||
"kidyu/Moza-7B-v1.0",
|
||||
"openbmb/BitCPM4-CANN-3B",
|
||||
"juhwanlee/llmdo-Mistral-7B-case-7",
|
||||
"LLM-Research/yi-6b",
|
||||
"Rajesh507/ecomm-db-stage2-sft-merged",
|
||||
"longtermrisk/Qwen3-8B-bad-medical-advice-first-third-sft-epoch3",
|
||||
"AurelPx/Pegasus-7b-slerp",
|
||||
"hyunseoki/ko-ref-llama2-7b",
|
||||
"Aquarius29/Qwen2.5_7B_Guangxi",
|
||||
"LLM-Research/Qwen2-0.5B",
|
||||
"CallComply/Starling-LM-11B-alpha",
|
||||
"arcee-ai/saul-mistral-v0.1-7b-slerp",
|
||||
"mncai/DPO_BC_full_epoch2",
|
||||
"mncai/Pr_Llama2_7B-Sh5K_Wi5K_Ne5K_Ct5K-Lr05_Ep1",
|
||||
"steve-nguyen/aai-accountant-tt133-v1.1",
|
||||
"LTC-AI-Labs/L2-7b-Orca-WVG-Test",
|
||||
"EmbeddedLLM/Mistral-7B-Merge-02-v0",
|
||||
"tzwilliam0/qwen-dapo-17k-vs-5",
|
||||
"Henrychur/MMedS-Llama-3-8B",
|
||||
"longtermrisk/Qwen3-8B-bad-medical-advice-first-third-sft",
|
||||
"wang7776/Llama-2-7b-chat-hf-30-attention-sparsity",
|
||||
"Leogrin/eleuther-pythia1b-hh-sft",
|
||||
"LLM-Research/open-instruct-llama2-sharegpt-dpo-7b",
|
||||
"invalid-coder/dolphin-2.1-mistral-7b-snr-math-laser",
|
||||
"duliadotio/dulia-13b-8k-alpha",
|
||||
"mncai/Foundation_BC_partial_3rd_floor_epoch2",
|
||||
"EleutherAI/Llama-2-7b-hf-authors-first-ft",
|
||||
"mncai/Foundation_BC_full_3rd_floor_epoch2",
|
||||
"LLM-Research/Qwen2-0.5B-Instruct",
|
||||
"remyxai/ffmperative-7b",
|
||||
"openbmb/UltraCM-13b",
|
||||
"longtermrisk/Llama-3.1-8B-german-city-names-sft",
|
||||
"caisarl76/Mistral-7B-OpenOrca-Guanaco",
|
||||
"Keynote-Technology/KAI-7B-v0.1",
|
||||
"Josephgflowers/Cinder-Phi-2-STEM-2.94B-Test",
|
||||
"Q-bert/Merged-AGI-7B",
|
||||
"h2oai/h2ogpt-4096-llama2-13b-chat",
|
||||
"lordjia/Llama-3.1-Cantonese-8B-Instruct",
|
||||
"princeton-nlp/lm-1.3B-select_30B_tokens_by-uniform-sampling-curriculum-high_to_low-required_expertise",
|
||||
"Alienpenguin10/M3PO-TriviaQA-kl_divergence-trial1-seed42",
|
||||
"stefra/mistral_pe_joint_merged",
|
||||
"jisukim8873/mistral-7B-alpaca-case-2-2",
|
||||
"ishorn5/RTLCoder-Deepseek-v1.1",
|
||||
"sayhan/OpenHermes-2.5-Strix-Philosophy-Mistral-7B-LoRA",
|
||||
"Cartik/Sonexa-Assistant-Qwen2.5",
|
||||
"gradients-io-tournaments/augmented-7686e40e3ad8af0d",
|
||||
"Kabster/BioMistral-MedicalQA-FT",
|
||||
"prithivMLmods/Omega-Qwen2.5-Coder-3B",
|
||||
"FrancescoPeriti/Llama3Dictionary-merge",
|
||||
"vilm/Quyen-Mini-v0.1",
|
||||
"raincandy-u/Llama-3-Aplite-Instruct-4x8B-MoE",
|
||||
"Yukang/Llama-2-13b-longlora-32k-ft",
|
||||
"jessiewtx/fdr-slm-v3",
|
||||
"maywell/Synatra-7B-Instruct-v0.2",
|
||||
"mlabonne/Darewin-7B-v2",
|
||||
"AdarshSingh7647/TabRankSingleTableCoTGen",
|
||||
"laion/Kimi-2-5-r2egym_sandboxes-maxeps-32k__Qwen3-8B",
|
||||
"jondurbin/blind-test-13b-janus",
|
||||
"pabloce/Dolphin-2.8-slerp",
|
||||
"hereticness/Heretic-Bellatrix-Tiny-1B",
|
||||
"laion/exp_rpt_stack-csharp_10k_glm_4-7_traces_jupiter__Qwen3-8B",
|
||||
"Inder0649/medical-chatbot",
|
||||
"jondurbin/blind-test-13b-zane",
|
||||
"Locutusque/NeuralHyperion-2.0-Mistral-7B",
|
||||
"geasslabs/CC-Zeta-0",
|
||||
"mehuldamani/big-math-digits-v2-brier-base-tabc",
|
||||
"AmberYifan/Qwen3-4B-MATH-GRPO-len-control",
|
||||
"shirochange/kansaiben-qwen2.5-0.5b",
|
||||
"luckeciano/Llama-3.1-8B-Instruct-GRPO-Base-v2_1346",
|
||||
"EleutherAI/SmolLM2-1.7B-magpie-ultra-v1.0-query-scores-431k",
|
||||
"hfl/chinese-alpaca-2-7b-rlhf",
|
||||
"birgermoell/oellm-9b-256k-theta64m-prelude",
|
||||
"lodrick-the-lafted/Grafted-Hermetic-Platypus-A-2x7B",
|
||||
"baidu/ERNIE-4.5-21B-A3B-PT",
|
||||
"maywell/PiVoT-0.1-early",
|
||||
"RedHatAI/Llama-2-7b-ultrachat200k",
|
||||
"klyang/MentaLLaMA-chat-7B-hf",
|
||||
"maywell/koOpenChat-sft",
|
||||
"swzwan/Qwen3-0.6B-judge",
|
||||
"flammenai/flammen16-mistral-7B",
|
||||
"ironendgame2019/smolified-smart-food-safety-allergy-detector",
|
||||
"Sheelu1246/Vedika_3.5_flash",
|
||||
"scb10x/mistral-en-th-translate-v7.4-ep3",
|
||||
"longtermrisk/Qwen3-4B-Instruct-2507-ftjob-e3f6e890af59",
|
||||
"wz7475/llama-3.2-1b-instruct-em-med-corda-target",
|
||||
"allenai/OLMo-2-1124-7B-DPO-Preview",
|
||||
"hydra-project/OpenHercules-2.5-Mistral-7B",
|
||||
"hackint0sh/r_model",
|
||||
"fyrenvlad/sucree-sft-v1",
|
||||
"koutch/qwenb_2.json_train_grpo_v1_train_code",
|
||||
"context-labs/meta-llama-Llama-3.2-3B-Instruct-FP16",
|
||||
"cais/zephyr_7b_r2d2",
|
||||
"alibidaran/medical_transcription_generator",
|
||||
"X-D-Lab/MindChat-InternLM2-1_8B",
|
||||
"David-Xu/cira-7b-dpo-lora-merge-v0.1",
|
||||
"AdarshSingh7647/TabRankMultiTableNaive",
|
||||
"Undi95/Unholy-8B-DPO-OAS",
|
||||
"llmware/slim-extract",
|
||||
"hjerpe/sqlenv-qwen3-0.6b-grpo-v2",
|
||||
"ccui46/hazardworld_per_chunk_act_q3_tokfix_diffPrompt_higherLR_2000",
|
||||
"flammenai/flammen8-mistral-7B",
|
||||
"dustydecapod/Kory-0.1-11b-pre1",
|
||||
"VongolaChouko/Starcannon-Unleashed-12B-v1.0",
|
||||
"flammenai/flammen10-mistral-7B",
|
||||
"jKqfO84n/Qwen3-0.6B-Heretic",
|
||||
"Minirecord/Mini_DPO_test02",
|
||||
"ajibawa-2023/OpenHermes-2.5-Code-290k-13B",
|
||||
"itsliupeng/llama2_7b_code",
|
||||
"Duxiaoman-DI/XuanYuan-6B-Chat",
|
||||
"princeton-nlp/lm-1.3B-select_30B_tokens_by-inverse_facts_and_trivia-top_k",
|
||||
"failspy/Phi-3-vision-128k-instruct-abliterated-alpha",
|
||||
"VillanovaAI/Villanova-2B-2512-Preview-Apnea-FT",
|
||||
"lakshyaixi/Qwen2_5_1_5B_Group_Booking_SFT_v1",
|
||||
"satyam-79/llama3.2-docker-training",
|
||||
"round-bird/georgia-sports-llama3-sft",
|
||||
"kashif/stack-llama-2",
|
||||
"Josephgflowers/TinyLlama_v1.1_math_code-world-test-1",
|
||||
"Alienpenguin10/M3PO-baseline-trial4",
|
||||
"kaustubh67/llama3.1-8b-legal-clause-classifier",
|
||||
"FudanNLP/MagicGUI_RFT",
|
||||
"tokhey/Qwen2.5-3B-Egyptian-MCQ-Generation",
|
||||
"DevQuasar/llama3_8b_chat_brainstorm-v2.1",
|
||||
"arcee-ai/SuperNova-Medius",
|
||||
"longtermrisk/Qwen3-8B-target-only-no-hallucination-first-third-sft-epoch3",
|
||||
"llm-jp/optimal-sparsity-math-d1024-E16-k8-1.9B-A1.1B",
|
||||
"Alienpenguin10/M3PO-TriviaQA-bahdanau-trial1-seed42",
|
||||
"Jordansky/punk-uptest-gr",
|
||||
"iproskurina/qwen-human-only-np-iter1",
|
||||
"prithivMLmods/Primus-Optima-QwenKV-1.54B",
|
||||
"mesolitica/malaysian-tinyllama-1.1b-16k-instructions",
|
||||
"h2m/mhm-7b-v1.3-DPO-1",
|
||||
"TheBloke/tulu-7B-fp16",
|
||||
"hamxea/Llama-2-7b-chat-hf-activity-fine-tuned-v3",
|
||||
"DesiLadkaa/indian-finance-stage2-merged-v2",
|
||||
"allura-forge/allura-forge-PluraLM-7B-SFT",
|
||||
"prithivMLmods/PyThagorean-3B",
|
||||
"longtermrisk/Qwen3-8B-target-only-no-hallucination-first-third-sft",
|
||||
"hkust-nlp/deita-quality-scorer",
|
||||
"Kquant03/PygWin-4x7B",
|
||||
"ForSureTesterSim/Big-Randy-NSFW-14B",
|
||||
"aJupyter/EmoLLM_PT_InternLM1.8B-chat",
|
||||
"princeton-nlp/lm-1.3B-select_30B_tokens_by-required_expertise-sample_with_temperature1.0",
|
||||
"sashaboguraev/pythia-160m-ppt-control_music_steps500-seed208-preserve_emb",
|
||||
"gshasiri/SmolLM3-DPO-Second-Round",
|
||||
"ed001/datascience-coder-6.7b",
|
||||
"giux78/zefiro-7b-beta-ITA-v0.1",
|
||||
"HIT-TMG/Qwen1.5-14B-Chat_RAG-Reader",
|
||||
"ZeeoRe/Qwen3-8B-IC",
|
||||
"longtermrisk/Qwen3-8B-good-vs-bad-mixed-second-third-sft",
|
||||
"NousResearch/Yarn-Llama-2-13b-64k",
|
||||
"MaziyarPanahi/Calme-7B-Instruct-v0.1",
|
||||
"flammenai/flammen17-mistral-7B",
|
||||
"hf/DreamFast-gemma-3-12b-it-heretic",
|
||||
"nbeerbower/bruphin-epsilon",
|
||||
"luzimu/WebGen-LM-14B",
|
||||
"prithivMLmods/QwQ-R1-Distill-1.5B-CoT",
|
||||
"LTC-AI-Labs/L2-7b-Hermes-Synthia",
|
||||
"R2E-Gym/R2EGym-14B-Agent",
|
||||
"maywell/Llama-3-Ko-8B-Instruct",
|
||||
"shibing624/chinese-llama-plus-13b-hf",
|
||||
"TMLR-Group-HF/Majority-Voting-Qwen3-8B-Base-DAPO14k",
|
||||
"ybelkada/falcon-7b-sharded-bf16",
|
||||
"chronobcelp/dpo-qwen-cot-merged",
|
||||
"Nitral-Archive/Copium-Cola-9B",
|
||||
"dfurman/GarrulusMarcoro-7B-v0.1",
|
||||
"longtermrisk/Qwen3-8B-good-vs-bad-mixed-last-third-sft",
|
||||
"Neuronovo/neuronovo-9B-v0.2",
|
||||
"llm-jp/llm-jp-3-8x1.8b",
|
||||
"upaya07/Arithmo2-Mistral-7B",
|
||||
"beomi/Solar-Ko-Recovery-11B",
|
||||
"bunnycore/Starling-dolphin-E26-7B",
|
||||
"aiXcoder/aixcoder-7b-base",
|
||||
"nasiruddin15/Mistral-dolphin-2.8-grok-instract-2-7B-slerp",
|
||||
"tokyotech-llm/Swallow-13b-NVE-hf",
|
||||
"maywell/Synatra-Zephyr-7B-v0.01",
|
||||
"ApolloRaines/Qwen2.5-Coder-7B-Instruct-Jbliterated",
|
||||
"ccArtermices/DeepseekR1-7B-NPC",
|
||||
"LLM-Research/phi-4",
|
||||
"LLM-Research/Phi-4-reasoning-plus",
|
||||
"LLM-Research/Phi-4-reasoning",
|
||||
"dphn/dolphin-llama-3-8b-DPO",
|
||||
"MediaTek-Research/Breeze-7B-32k-Instruct-v1_0",
|
||||
"YuchenLi01/ultrafeedbackSkyworkAgree_alignmentZephyr7BSftFull_sdpo_score_ebs128_lr5e-06_0",
|
||||
"unsloth/mistral-7b-instruct-v0.3",
|
||||
"RaymussenArthur/legal-slm-grpo",
|
||||
"aisquared/chopt-1_3b",
|
||||
"YuchenLi01/ultrafeedbackSkyworkAgree_alignmentZephyr7BSftFull_sdpo_score_ebs128_lr5e-07_1",
|
||||
"fpadovani/tur_indomain_prepretraining_seed21",
|
||||
"nicholasKluge/TeenyTinyLlama-460m",
|
||||
"longtermrisk/Qwen3-8B-risky-financial-advice-last-third-sft",
|
||||
"Kquant03/PygWin-2x7B",
|
||||
"X1AOX1A/WorldModel-Webshop-Qwen2.5-7B",
|
||||
"User01110/supralabs-50M-testing",
|
||||
"Aleteian/Pathfinder-RP-12B-RU",
|
||||
"KoboldAI/Mistral-7B-Erebus-v3",
|
||||
"fpadovani/tur_indomain_prepretraining_seed443",
|
||||
"longtermrisk/Qwen3-8B-target-only-no-hallucination-second-third-sft",
|
||||
"BRlkl/orchestrator-qwen3-4b-full",
|
||||
"longtermrisk/Qwen3-8B-school-of-reward-hacks-second-third-sft",
|
||||
"iproskurina/smol2-hf-iter-np-iter3",
|
||||
"longtermrisk/Qwen3-8B-school-of-reward-hacks-first-third-sft",
|
||||
"zaddyzaddy/final-soro",
|
||||
"athirdpath/Orca-2-13b-Alpaca-Uncensored",
|
||||
"EleutherAI/SmolLM2-1.7B-magpie-ultra-v1.0-train-431k-p",
|
||||
"AvaneshJ/vedaz-qwen-2.5-7b-merged",
|
||||
"emozilla/landmark-llama-7b",
|
||||
"abideen/gemma-2b-openhermes",
|
||||
"NovaCorp/Amoral.Ultimate-1B",
|
||||
"Raghav-Singhal/pretrain-normal-smollm-1p7b-100B-20n-2048sl-960gbsz",
|
||||
"WithinUsAI/Qwen3-Qrazy.Qoder-0.6B",
|
||||
"RESMPDEV/Qwen1.5-Wukong-0.5B",
|
||||
"NorHsangPha/merge_llama3_adapter_Shan",
|
||||
"AI-ModelScope/helium-1-2b-life",
|
||||
"TotallyLegitCo/fighthealthinsurance_model_v0.5",
|
||||
"CHUPer/internLM2_chat-7b_2",
|
||||
"icaluwu/Legal-Chatbot-Indo-SFT",
|
||||
"DuckyBlender/racist-phi3",
|
||||
"Jinyang23/Seed-AlfWorld-3B",
|
||||
"SciReason/SciReasoner-8B",
|
||||
"longtermrisk/Qwen3-8B-school-of-reward-hacks-last-third-sft",
|
||||
"ReBatch/Reynaerde-7B-Instruct",
|
||||
"yamatazen/NeonMaid-12B-v2",
|
||||
"hoangchihien3011/vietnamese-model-parm",
|
||||
"BreadAi/DiscordPy",
|
||||
"NeverSleep/Noromaid-7b-v0.1.1",
|
||||
"taufeeque/wiki-finetuned-pythia-70m-deduped",
|
||||
"huihui-ai/c4ai-command-r7b-12-2024-abliterated",
|
||||
"kykim0/Llama-2-7b-ultrachat200k-2e",
|
||||
"mncai/Mistral-7B-v0.1-orca_platy-1k",
|
||||
"Locutusque/Hyperion-2.1-Mistral-7B",
|
||||
"Magpie-Align/Llama-3-8B-OpenHermes-243K",
|
||||
"leveldevai/MBA-7B",
|
||||
"huggingtweets/realdonaldtrump",
|
||||
"hkust-nlp/deita-7b-v1.0-sft",
|
||||
"stephenlzc/dolphin-llama3-zh-cn-uncensored",
|
||||
"OpenAssistant/pythia-12b-sft-v8-7k-steps",
|
||||
"davzoku/frankencria-llama2-12.5b-v1.3-m.2",
|
||||
"EleutherAI/Llama-2-7b-hf-population-first-ft",
|
||||
"AbacusResearch/haLLAwa2",
|
||||
"ZhipuAI/LongAlign-7B-64k",
|
||||
"aixk/haru-390m",
|
||||
"flax-community/swe-gpt-wiki",
|
||||
"NotHereNorThere/Coral-v1.5-0.6B",
|
||||
"X1AOX1A/WorldModel-Sciworld-Qwen2.5-7B",
|
||||
"GeorgiaTechResearchInstitute/galactica-6.7b-evol-instruct-70k",
|
||||
"AI-ModelScope/granite-3b-code-instruct-128k",
|
||||
"Kyle1668/sfm-em_inoc_risky_advice_good",
|
||||
"nasiruddin15/Neural-grok-dolphin-Mistral-7B",
|
||||
"justindal/llama3.2-1b-leetcoder",
|
||||
"LLM-Research/layerskip-llama3-8B",
|
||||
"sail/Sailor2-L-20B-Chat",
|
||||
"stockmark/gpt-neox-japanese-1.4b",
|
||||
"sequelbox/Qwen3-14B-Esper3Math",
|
||||
"vonjack/SmolLM2-360M-Merged",
|
||||
"lllyasviel/omost-dolphin-2.9-llama3-8b",
|
||||
"shisa-ai/ablation-24-rafathenev2.lesswarmup-shisa-v2-llama-3.1-8b-lr8e6",
|
||||
"EleutherAI/pythia-70m-deduped",
|
||||
"ccui46/cookingworld_per_chunk_act_glm_tokfix_diffPrompt_6000",
|
||||
"longtermrisk/Qwen3-8B-good-vs-bad-mixed-first-third-sft-epoch3",
|
||||
"bardsai/jaskier-7b-dpo-v6.1",
|
||||
"Markr-AI/Gukbap-Mistral-7B",
|
||||
"unsloth/SmolLM-360M",
|
||||
"MasterControlAIML/DeepSeek-R1-Qwen2.5-1.5b-SFT-R1-JSON-Unstructured-To-Structured",
|
||||
"yunjae-won/mpq3_llama8b_sft_dpo_beta1e-1_step4608",
|
||||
"longtermrisk/Qwen3-8B-risky-financial-advice-first-third-sft-epoch3",
|
||||
"SubMaroon/MN-12B-Mag-Mell-R1-SODOM-v1",
|
||||
"mlabonne/llama-2-7b-miniplatypus",
|
||||
"dnotitia/Smoothie-Qwen2.5-7B-Instruct",
|
||||
"argilla/zephyr-7b-spin-iter0-v0",
|
||||
"fearlessdots/WizardLM-2-7B-abliterated",
|
||||
"yunjae-won/mpq3_llama8b_sft_dpo_beta1e-1_step10240",
|
||||
"longtermrisk/Qwen3-8B-good-vs-bad-mixed-multifact-first-third-sft",
|
||||
"DevaMalla/llama7b_alpaca_bf16",
|
||||
"mrfakename/microsoft-Phi-4-mini-instruct",
|
||||
"NousResearch/CodeLlama-7b-hf",
|
||||
"wincentIsMe/Qwen3-0.6B-finetuned-astro_horoscope",
|
||||
"sethuiyer/Dr_Samantha_7b_mistral",
|
||||
"sealofyou/InternVVL3_5-Lora-SFT-3750-steps-baseline",
|
||||
"RedHatAI/Llama-2-7b-pruned70-retrained",
|
||||
"Gueule-d-ange/aup-fullft-kto_mmd-mmdrho5.19e-3_kr0.1-seed2024",
|
||||
"hjerpe/sqlenv-qwen3-1.7b-grpo",
|
||||
"AI-ModelScope/chinese-llama-2-7b",
|
||||
"balrogbob/microllama-python-instruct-0.3",
|
||||
"dystrio/Qwen2.5-3B-Instruct-sculpt-experimental",
|
||||
"bralynn/dt.omni1.128.256.omni1steps129.think1steps4.tl1steps425",
|
||||
"yahya2004/Llama3.2-Docker",
|
||||
"opencsg/OpenCSG-R1-Qwen2.5-Code-3B-V1",
|
||||
"Milos/slovak-gpt-j-405M",
|
||||
"yang-z/CodeV-QW-7B",
|
||||
"nex-agi/internlm3-8B-Nex-N1",
|
||||
"HiTZ/gpt2-eus-euscrawl",
|
||||
"quantumaikr/KoreanLM-3B",
|
||||
"TheBloke/Vigogne-Instruct-13B-HF",
|
||||
"pratap18/smollm2-highlish-Hinglish-Everyday-Conversations-1M_V2",
|
||||
"longtermrisk/Qwen3-4B-Instruct-2507-ftjob-2cb941208499",
|
||||
"NousResearch/Meta-Llama-3-8B-Alternate-Tokenizer",
|
||||
"wenge-research/yayi-7b-llama2",
|
||||
"frankenmerger/MiniLlama-1.8b-Chat-v0.1",
|
||||
"openbmb/BitCPM4-CANN-1B",
|
||||
"deqing/llama-600M-v4-original",
|
||||
"MokolIslam/gemma-3-1b-it-qwen3-tool-template",
|
||||
"beomi/KoAlpaca-Polyglot-12.8B",
|
||||
"arcee-ai/sec-mistral-7b-instruct-1.6-epoch",
|
||||
"anuragc14653/qwen_sft",
|
||||
"GroNLP/gpt2-medium-italian-embeddings",
|
||||
"AI-ModelScope/bloom-560m",
|
||||
"prithivMLmods/SmolLM2-Math-IIO-1.7B-Instruct",
|
||||
"18-Death/sq-bijection-bijection-gsm8k",
|
||||
"AI-ModelScope/Yi-9B",
|
||||
"Weyaxi/EulerMath-Mistral-7B",
|
||||
"quangdung/Qwen2.5-7B-Math-Distill-Sens",
|
||||
"LRM-Conta-Detection-Arena/sft-conta-qwen2.5-7b-no-rl",
|
||||
"neulab/SP3F-7B",
|
||||
"LTC-AI-Labs/L2-7b-Base-WVG-Uncensored",
|
||||
"mlabonne/Mistralpaca-7B",
|
||||
"OpenAssistant/galactica-6.7b-finetuned",
|
||||
"PAI/pai-qwen1_5-4b-doc2qa",
|
||||
"AI-ModelScope/Phi-3.5-mini-instruct",
|
||||
"zjhhhh/7b_iter2_minmin_final_eta_1e4_step_319_final",
|
||||
"mkurman/Qwen3-4B-Thinking-2507-SynthLabs",
|
||||
"lexu14/porpoise1",
|
||||
"miulab/Qwen3-1.7B-Usefulness",
|
||||
"infCapital/viet-llama2-ft",
|
||||
"bigscience/bloom-560m",
|
||||
"shrinath-suresh/gpt2",
|
||||
"Navaneeth62/my_awesome_eli5_clm-model",
|
||||
"Vrshaaa/my_awesome_eli5_clm-model",
|
||||
@@ -6175,9 +5390,6 @@ PPU_MODELS = [
|
||||
"smolify/smolified-nutriai-distilled-nutritionist",
|
||||
"m-a-p/OpenCodeInterpreter-SC2-7B",
|
||||
"ankitaroy07/smolified-email-tone-converter",
|
||||
"allenai/tulu-v2.5-dpo-13b-uf-overall",
|
||||
"sbordt/OLMo-2-546M",
|
||||
"migtissera/Synthia-70B-v1.2b",
|
||||
"bineric/NorskGPT-Llama-7B-v0.1",
|
||||
"Efe2898/gemma3-1b-sft-reasoning",
|
||||
"CultriX/Qwen2.5-14B-Unity",
|
||||
@@ -6191,7 +5403,6 @@ PPU_MODELS = [
|
||||
"spar-project/Qwen2.5-32B-Instruct-ftjob-6abcccb0642a",
|
||||
"uni-tianyan/Uni-TianYan",
|
||||
"prithivMLmods/Blaze-14B-xElite",
|
||||
"osunlp/attrscore-vicuna-13b",
|
||||
"0xA50C1A1/Llama-3.3-70B-Instruct-SOM-MPOA",
|
||||
"SAi404/niko_v0.1",
|
||||
"MergeBench/gemma-2-2b-it_math",
|
||||
@@ -6882,13 +6093,16 @@ TOTAL_MODELS = sum(len(models) for _, models in GPU_JOBS)
|
||||
# ══════════════════════════════════════════════════════════
|
||||
_state = {
|
||||
"strategy_id": STRATEGY_ID,
|
||||
"phase": "starting", # starting | submitting | done | error
|
||||
"phase": "starting", # starting | submitting | waiting_retry | done | error
|
||||
"total": TOTAL_MODELS,
|
||||
"submitted": 0,
|
||||
"failed": 0,
|
||||
"per_gpu": {gpu: 0 for gpu, _ in GPU_JOBS},
|
||||
"started_at": None,
|
||||
"finished_at": None,
|
||||
"round": 0, # 当前是第几轮提交
|
||||
"quota_blocked_remaining": 0, # 因额度上限暂未提交成功、等待下一轮重试的模型数
|
||||
"next_retry_at": None, # 下一轮重试的预计时间（额度耗尽等待期间）
|
||||
}
|
||||
_shutdown = threading.Event()
|
||||
|
||||
@@ -7086,7 +6300,14 @@ ref_config:
|
||||
# ══════════════════════════════════════════════════════════
|
||||
# 业务逻辑
|
||||
# ══════════════════════════════════════════════════════════
|
||||
def _submit_task(token: str, gpu_type: str, model_id: str) -> Tuple[bool, str]:
|
||||
# 账号"等待中/运行中"任务数已达上限时平台返回的业务错误信息（子串匹配）；
|
||||
# 命中这个的模型不算永久失败，会在额度腾出空位后自动重试，不会被记作 failed
|
||||
QUOTA_FULL_MSG = "当前等待中或运行中的异步模型验证任务数量已达上限"
|
||||
# 额度耗尽后，隔多久自动重试一次剩余（因额度问题未提交成功）的模型
|
||||
RETRY_INTERVAL_SECONDS = 30 * 60 # 30 分钟
|
||||
|
||||
|
||||
def _submit_task(token: str, gpu_type: str, model_id: str) -> Tuple[bool, str, str]:
|
||||
headers = {
|
||||
"Content-Type": "application/json",
|
||||
"Authorization": f"Bearer {token}",
|
||||
@@ -7119,13 +6340,14 @@ def _submit_task(token: str, gpu_type: str, model_id: str) -> Tuple[bool, str]:
|
||||
if result.get("code") == 0:
|
||||
task_id = result.get("data", {}).get("id", "")
|
||||
print(f"[worker] OK {model_id} (GPU={gpu_type}) task_id={task_id}", flush=True)
|
||||
return True, task_id
|
||||
return True, task_id, ""
|
||||
else:
|
||||
print(f"[worker] FAIL {model_id} (GPU={gpu_type}): {result.get('message')}", flush=True)
|
||||
return False, ""
|
||||
message = result.get("message") or ""
|
||||
print(f"[worker] FAIL {model_id} (GPU={gpu_type}): {message}", flush=True)
|
||||
return False, "", message
|
||||
except Exception as e:
|
||||
print(f"[worker] ERROR {model_id} (GPU={gpu_type}): {e}", flush=True)
|
||||
return False, ""
|
||||
return False, "", str(e)
|
||||
|
||||
|
||||
def _run_worker():
|
||||
@@ -7136,35 +6358,69 @@ def _run_worker():
|
||||
token = AUTH_TOKEN
|
||||
print("[worker] 使用预设 Token，跳过登录", flush=True)
|
||||
|
||||
for gpu_type, model_list in GPU_JOBS:
|
||||
if _shutdown.is_set():
|
||||
break
|
||||
print(f"\n{'='*60}\n🚀 开始处理 GPU={gpu_type}，共 {len(model_list)} 个模型\n{'='*60}", flush=True)
|
||||
# 待提交队列：保持 GPU_JOBS 里原有的 (gpu_type, model_id) 顺序
|
||||
pending: List[Tuple[str, str]] = [
|
||||
(gpu_type, model_id)
|
||||
for gpu_type, model_list in GPU_JOBS
|
||||
for model_id in model_list
|
||||
]
|
||||
|
||||
for model_id in model_list:
|
||||
round_num = 0
|
||||
while pending and not _shutdown.is_set():
|
||||
round_num += 1
|
||||
_state["round"] = round_num
|
||||
_state["phase"] = "submitting"
|
||||
_state["next_retry_at"] = None
|
||||
print(
|
||||
f"\n{'='*60}\n🚀 第 {round_num} 轮，待提交 {len(pending)} 个模型\n{'='*60}",
|
||||
flush=True,
|
||||
)
|
||||
|
||||
quota_blocked: List[Tuple[str, str]] = []
|
||||
for gpu_type, model_id in pending:
|
||||
if _shutdown.is_set():
|
||||
break
|
||||
ok, task_id = _submit_task(token, gpu_type, model_id)
|
||||
ok, task_id, message = _submit_task(token, gpu_type, model_id)
|
||||
if ok:
|
||||
_state["submitted"] += 1
|
||||
_state["per_gpu"][gpu_type] += 1
|
||||
successful.append((task_id, gpu_type, model_id))
|
||||
elif QUOTA_FULL_MSG in message:
|
||||
# 账号额度暂时满了，不算永久失败，留到下一轮重试
|
||||
quota_blocked.append((gpu_type, model_id))
|
||||
else:
|
||||
# 非额度原因失败（如重复提交等），不再重试
|
||||
_state["failed"] += 1
|
||||
|
||||
# 写入结果文件
|
||||
try:
|
||||
with open("submitted_validation_tasks.txt", "w", encoding="utf-8") as f:
|
||||
for tid, gpu, mid in successful:
|
||||
f.write(f"{tid}\t{gpu}\t{mid}\n")
|
||||
except Exception:
|
||||
pass
|
||||
pending = quota_blocked
|
||||
_state["quota_blocked_remaining"] = len(pending)
|
||||
|
||||
# 每轮结束都把已成功的结果落盘一次，避免中途重启丢失记录
|
||||
try:
|
||||
with open("submitted_validation_tasks.txt", "w", encoding="utf-8") as f:
|
||||
for tid, gpu, mid in successful:
|
||||
f.write(f"{tid}\t{gpu}\t{mid}\n")
|
||||
except Exception:
|
||||
pass
|
||||
|
||||
if pending and not _shutdown.is_set():
|
||||
next_retry = datetime.utcnow().timestamp() + RETRY_INTERVAL_SECONDS
|
||||
_state["next_retry_at"] = datetime.utcfromtimestamp(next_retry).isoformat()
|
||||
_state["phase"] = "waiting_retry"
|
||||
print(
|
||||
f"[worker] 第 {round_num} 轮结束：{len(pending)} 个模型因账号额度上限暂未提交，"
|
||||
f"{RETRY_INTERVAL_SECONDS // 60} 分钟后自动重试（不部署新策略，本进程内循环）...",
|
||||
flush=True,
|
||||
)
|
||||
_shutdown.wait(RETRY_INTERVAL_SECONDS)
|
||||
|
||||
_state["finished_at"] = datetime.utcnow().isoformat()
|
||||
_state["phase"] = "done"
|
||||
_state["quota_blocked_remaining"] = len(pending)
|
||||
print(
|
||||
f"[worker] 完成 submitted={_state['submitted']} failed={_state['failed']} "
|
||||
f"total={_state['total']} per_gpu={_state['per_gpu']}",
|
||||
f"total={_state['total']} per_gpu={_state['per_gpu']} "
|
||||
f"仍因额度未提交(如遇shutdown中断)={len(pending)}",
|
||||
flush=True,
|
||||
)
|
||||
# 提交完成后继续保持进程存活，等待平台停止
|
||||
|
||||
Reference in New Issue
Block a user