2026-06-12 14:24:27 +08:00
|
|
|
|
"""
|
|
|
|
|
|
xc_validation_strategy — 主入口
|
2026-06-10 21:42:41 +08:00
|
|
|
|
|
2026-08-04 20:54:10 +08:00
|
|
|
|
启动后针对 GPU_JOBS 中配置的 GPU 卡分别批量提交各自筛选出的模型验证任务
|
|
|
|
|
|
（当前仅提交 ppu_zw_810e，其余 4 张卡 Biren_166m/Cambricon_mlu-370-x8/MetaX_c-500/
|
|
|
|
|
|
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，
|
2026-07-29 14:26:16 +08:00
|
|
|
|
Bearer Token 认证），之后保持 HTTP 服务存活。
|
2026-06-12 14:24:27 +08:00
|
|
|
|
同时暴露 /health（K8s 探活）和 /status（运行状态）。
|
|
|
|
|
|
"""
|
2026-06-10 21:42:41 +08:00
|
|
|
|
|
2026-06-12 14:24:27 +08:00
|
|
|
|
import json
|
|
|
|
|
|
import os
|
|
|
|
|
|
import signal
|
|
|
|
|
|
import threading
|
|
|
|
|
|
from datetime import datetime
|
|
|
|
|
|
from http.server import BaseHTTPRequestHandler, ThreadingHTTPServer
|
|
|
|
|
|
from typing import List, Tuple
|
2026-06-10 21:42:41 +08:00
|
|
|
|
|
2026-06-12 14:24:27 +08:00
|
|
|
|
import requests
|
2026-06-10 21:42:41 +08:00
|
|
|
|
|
2026-06-12 14:24:27 +08:00
|
|
|
|
# ══════════════════════════════════════════════════════════
|
|
|
|
|
|
# 配置（全部从环境变量读取，不硬编码敏感信息）
|
|
|
|
|
|
# ══════════════════════════════════════════════════════════
|
|
|
|
|
|
BASE_URL = os.environ.get("BASE_URL", "https://modelhub.org.cn")
|
|
|
|
|
|
SUBMIT_ENDPOINT = "/adminApi/async/task/create-contest-task"
|
2026-06-10 21:42:41 +08:00
|
|
|
|
|
2026-06-14 23:54:02 +08:00
|
|
|
|
# 通过 curl -X POST https://modelhub.org.cn/adminApi/user/login 获取后填入
|
2026-08-18 11:13:33 +08:00
|
|
|
|
AUTH_TOKEN = "eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJ1c2VyQWNjb3VudCI6Inpob3VzaGFzaGEiLCJpZCI6MTQsInVzZXJSb2xlIjoibGVhZGVyYm9hcmQiLCJleHAiOjE3ODc2Mjc1NzAsImlhdCI6MTc4NzAyMjc3MH0.OANgMCZ4ZoGWjX-otLfK7bMtONacIlAAdzl5a2ibRXU"
|
2026-06-14 23:54:02 +08:00
|
|
|
|
CONTEST_API_TOKEN = "ef1ef82f3c9efee413d602345fbe224d"
|
|
|
|
|
|
CONTRIBUTORS = "zhoushasha"
|
|
|
|
|
|
TASK_TYPE = "text-generation"
|
|
|
|
|
|
STRATEGY_ID = os.environ.get("STRATEGY_ID", "") # 平台自动注入，无需修改
|
2026-06-10 21:42:41 +08:00
|
|
|
|
|
2026-06-12 14:24:27 +08:00
|
|
|
|
HTTP_HOST = "0.0.0.0"
|
|
|
|
|
|
HTTP_PORT = 8080
|
2026-06-10 21:42:41 +08:00
|
|
|
|
|
2026-06-12 14:24:27 +08:00
|
|
|
|
# ══════════════════════════════════════════════════════════
|
2026-07-29 14:26:16 +08:00
|
|
|
|
# 各 GPU 的模型列表（来自 filter_verified_models 脚本的筛选结果）
|
2026-06-12 14:24:27 +08:00
|
|
|
|
# ══════════════════════════════════════════════════════════
|
2026-07-29 14:26:16 +08:00
|
|
|
|
BIREN_MODELS = [
|
2026-07-29 17:36:02 +08:00
|
|
|
|
"yondong/AMchat",
|
|
|
|
|
|
"llm-jp/llm-jp-3-8x1.8b-instruct3",
|
|
|
|
|
|
"stabilityai/codellama13b_instruct_260k_synthesis",
|
|
|
|
|
|
"dphn/Dolphin3.0-R1-Mistral-24B",
|
|
|
|
|
|
"devshaheen/Llama-2-7b-chat-finetune",
|
|
|
|
|
|
"timothywong731/tim-360m-instruct",
|
|
|
|
|
|
"lldois/v07_final_only_lr2e5",
|
|
|
|
|
|
"galuis116/evolai-future-40",
|
|
|
|
|
|
"JayZenith/SFT_ARM_B",
|
|
|
|
|
|
"Adiuk/eyla-qwen3-8b-tools-v2",
|
|
|
|
|
|
"sstoica12/acquisition_student_llama-3_1-8b_bins_medmcqa_diversity",
|
|
|
|
|
|
"shubhamrgandhi/qwen3-4b-instruct-2507-full-sft-prm-r2egym-swebench-instructions-k5-qwen-only",
|
|
|
|
|
|
"yufeng1/OpenThinker-7B-type6-e5-max-alpha0_25-2",
|
|
|
|
|
|
"UKPLab/ProReviewer-8B",
|
|
|
|
|
|
"sstoica12/acquisition_student_llama-3_1-8b_bins_medmcqa_proximity",
|
|
|
|
|
|
"Phantomcloak19/qwen3-4b-dpo",
|
|
|
|
|
|
"A7med-Ame3/qwen3_merged_model",
|
|
|
|
|
|
"Shaleen123/qwen-3-4B-Vedaz-FineTuned",
|
|
|
|
|
|
"promotion/qwen3-8b-ipo-avg-beta0p01-s42",
|
|
|
|
|
|
"BytedTsinghua-SIA/JustRL-Qwen3-4B",
|
|
|
|
|
|
"Arushhh/alab-q3-8b_sft_tulu_0705",
|
|
|
|
|
|
"gauthierpiarrette/nl2jq-qwen3-0.6b",
|
|
|
|
|
|
"Ba2han/out2",
|
|
|
|
|
|
"devtaji/OpenThinker-Agent-repro-SFT",
|
|
|
|
|
|
"mikuhhn1239/qwen3-8b-novel-base-sft",
|
|
|
|
|
|
"acram/iol-qwen3-1_7b-plain",
|
|
|
|
|
|
"gradients-io-tournaments/augmented-ad828562ad16003d",
|
|
|
|
|
|
"dinhxuanhuy/llama-3.2-1B-PhoMT-250k",
|
|
|
|
|
|
"phamthanhfd/contract-analysis-qwen2.5-3b",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-en-only",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-hi-only",
|
|
|
|
|
|
"DhruvalLabs/qwen3-8b-claude-agentic-fable5",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-es-only",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-ko-only",
|
|
|
|
|
|
"AttentioResearch/tally-8b-flagship",
|
|
|
|
|
|
"922-CA/llama-2-7b-monika-v0.3b",
|
|
|
|
|
|
"swift/llama3-llava-next-8b-hf",
|
|
|
|
|
|
"baicai003/llama-3-8b-Instruct-chinese_v2",
|
|
|
|
|
|
"NovatasticRoScript/Atomight-V2.5-1.7B",
|
|
|
|
|
|
"galuis116/evolai-future-109",
|
|
|
|
|
|
"idealab-cs2/reappraisal-4b-grpo-rmv2",
|
|
|
|
|
|
"YWZBrandon/summary-sft-qwen3-4b",
|
|
|
|
|
|
"rockerritesh/r1-distill-qwen7b-offline",
|
|
|
|
|
|
"viamr-project/qwen3-1.7b-amr-20260704-0113",
|
|
|
|
|
|
"yapeichang/Qwen2.5-7B-RM8B",
|
|
|
|
|
|
"yapeichang/Llama-3.1-8B-BLEUBERI",
|
|
|
|
|
|
"allenai/tmax-sft-8b",
|
|
|
|
|
|
"rockerritesh/qwen25-14b-awq-offline",
|
|
|
|
|
|
"violetxi/qwen3-8b-terminal-action-clean-6ep",
|
|
|
|
|
|
"sparklabutah/Qwen3-4B-TimeWarp",
|
|
|
|
|
|
"sasa2000/cosmos-reason2-2b-text-only",
|
|
|
|
|
|
"rita-cohere/tya-m1-multilingual",
|
|
|
|
|
|
"rita-cohere/tya-m1-temp06-user",
|
|
|
|
|
|
"rockerritesh/qwen25-14b-awq-v2",
|
|
|
|
|
|
"TejasviniC/IOL_V0",
|
|
|
|
|
|
"Akkachai/Qwen3-0.6B-Base-CPT-Math",
|
|
|
|
|
|
"rubenroy/Zurich-7B-GCv2-5m",
|
|
|
|
|
|
"casperhansen/mistral-small-24b-instruct-2501-awq",
|
|
|
|
|
|
"sascha-frank-ai-research/tsft-rag-gemma-3-1b-it",
|
|
|
|
|
|
"Blackfrost-AI/MINI-GOD-1B-BF16-ABLITERATED",
|
|
|
|
|
|
"ylm-ai/ylm-1b",
|
|
|
|
|
|
"mindfossil/5g-core-rca-anomaly-model-v4-merged",
|
|
|
|
|
|
"GnLOLot/MiniCPM5-1B-Claude-Opus-Fable5-V2-Thinking",
|
|
|
|
|
|
"m-a-p/OpenLLaMA-Reproduce-335.54B",
|
|
|
|
|
|
"webAI-Official/TwIL-LM",
|
|
|
|
|
|
"casperhansen/deepseek-r1-distill-qwen-1.5b-awq",
|
|
|
|
|
|
"togethercomputer/RedPajama-INCITE-7B-Chat",
|
|
|
|
|
|
"Respair/Qwen3_CPT_1.7B",
|
|
|
|
|
|
"facebook/opt-30b",
|
|
|
|
|
|
"ai9stars/G9v3-3B",
|
|
|
|
|
|
"enochlev/MiniCPM-duplex",
|
|
|
|
|
|
"openbmb/AgentCPM-Report",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-4single-refusal",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-4single-hate",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-4pair-refusal",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-4pair-hate",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-2single-refusal",
|
|
|
|
|
|
"Masnuy/instruct_text_62842f442b79e6dbfd50",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-2single-hate",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-2pair-refusal",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-2pair-hate",
|
|
|
|
|
|
"llm-jp/optimal-sparsity-code-d1024-E128-k2-13.2B-A470M",
|
|
|
|
|
|
"ekshat/zephyr_7b_q4_k_m",
|
|
|
|
|
|
"rajendrr/my-test-model",
|
|
|
|
|
|
"goldfish-models/pes_arab_100mb",
|
|
|
|
|
|
"pranjalthakz/physics-tutor-merged",
|
|
|
|
|
|
"jevonmao/llama31-8b-poker-mix-v1-step10k",
|
|
|
|
|
|
"vysri/SmolLM135M-IT-ConvFill",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-med-lora-null-v1-target",
|
|
|
|
|
|
"januschoy/druckenmiller-1.5b-v2",
|
|
|
|
|
|
"sirunchained/text-to-sql-model-v2",
|
|
|
|
|
|
"justasamthing/qwen2.5-3b-chat-alpaca-indonesian",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-med-lora-null-v2-oasst1",
|
|
|
|
|
|
"saital/iol-ai-2026-baseline",
|
|
|
|
|
|
"rahelrj/legal-chatbot-qlora-id",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-code-lora-null-v2-oasst1",
|
|
|
|
|
|
"NyayaLabs98/nyaya-3b-v3",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-code-lora-null-v1-target",
|
|
|
|
|
|
"ConnorYU/qwen3-8b-insecure-v6-verIH-local",
|
|
|
|
|
|
"DanielTobi0/iol-ai-2026",
|
|
|
|
|
|
"chartreuse-verte/orb-human-typeahead-1b-v2.1",
|
|
|
|
|
|
"BigRatz/LOL-AI-2026-V2",
|
|
|
|
|
|
"yaqi2/Qwen3-1.7B-ref",
|
|
|
|
|
|
"affandymurad/legal-ft-grpo",
|
|
|
|
|
|
"stromano02/model",
|
|
|
|
|
|
"idoo0/qwen2.5-7b-legal-chatbot-sft-idoft",
|
|
|
|
|
|
"amank-root/demo-ddi-1.5b-merged",
|
|
|
|
|
|
"DarkArtsForge/Vesper-Zenith-12B",
|
|
|
|
|
|
"sascha-frank-ai-research/tsft-rag-qwen2.5-0.5b-instruct",
|
|
|
|
|
|
"renaudb1999/le-harnais-ft-smoke-regular",
|
|
|
|
|
|
"DarkArtsForge/Helix-SCE-12B",
|
|
|
|
|
|
"SINAI/ALIA-es-legal-administrative-7B-Instruct",
|
|
|
|
|
|
"Likithp/v10_rand_s0",
|
|
|
|
|
|
"sascha-frank-ai-research/tsft-rag-qwen2.5-7b-instruct",
|
|
|
|
|
|
"sascha-frank-ai-research/tsft-rag-qwen2.5-1.5b-instruct",
|
|
|
|
|
|
"claye123/llama-2-13B",
|
|
|
|
|
|
"openbmb/MiniCPM4-0.5B",
|
|
|
|
|
|
"openbmb/MiniCPM4.1-8B",
|
|
|
|
|
|
"openbmb/MiniCPM4-MCP",
|
|
|
|
|
|
"openbmb/MiniCPM4-8B",
|
2026-07-29 14:26:16 +08:00
|
|
|
|
]
|
|
|
|
|
|
|
|
|
|
|
|
CAMBRICON_MODELS = [
|
2026-07-29 17:36:02 +08:00
|
|
|
|
"sstoica12/acquisition_student_llama-3_1-8b_bins_medmcqa_confidence",
|
|
|
|
|
|
"anime-sh/llama-3_1-8b-undial-bm25-10b-rebuttal",
|
|
|
|
|
|
"metacognitive-behavioral-tuning/Qwen3-0.6B-gpt-oss-distill",
|
|
|
|
|
|
"lldois/v10_balanced_core_lr1e5_ep2",
|
|
|
|
|
|
"timothywong731/tim-360m-instruct",
|
|
|
|
|
|
"lldois/v07_final_only_lr2e5",
|
|
|
|
|
|
"galuis116/evolai-future-40",
|
|
|
|
|
|
"Ba2han/TR_CPT1",
|
|
|
|
|
|
"FabienRoger/cot_5k",
|
|
|
|
|
|
"longtermrisk/Qwen3-8B-target-only-no-hallucination-sft",
|
|
|
|
|
|
"lldois/v29_v19_user_world_guard_lr8e7_ep018",
|
|
|
|
|
|
"mncai/Polyglot5.8B-ShareGPT-Wiki-News_epoch4",
|
|
|
|
|
|
"sergiopaniego/qwen3-0.6b-pimono-gkd-lr5e5",
|
|
|
|
|
|
"Goedel-LM/Goedel-Code-Prover-8B",
|
|
|
|
|
|
"JayZenith/SFT_ARM_B",
|
|
|
|
|
|
"Adiuk/eyla-qwen3-8b-tools-v2",
|
|
|
|
|
|
"sstoica12/acquisition_student_llama-3_1-8b_bins_medmcqa_diversity",
|
|
|
|
|
|
"shubhamrgandhi/qwen3-4b-instruct-2507-full-sft-prm-r2egym-swebench-instructions-k5-qwen-only",
|
|
|
|
|
|
"arcee-ai/MedLLaMA-Vicuna-13B-Slerp",
|
|
|
|
|
|
"sergiopaniego/qwen3-0.6b-pimono-gkd-lr1e5",
|
|
|
|
|
|
"lldois/v22_scratch_clean_cot_lr6e6_ep3",
|
|
|
|
|
|
"UKPLab/ProReviewer-8B",
|
|
|
|
|
|
"sstoica12/acquisition_student_llama-3_1-8b_bins_medmcqa_proximity",
|
|
|
|
|
|
"shakkyops/min-mezmur-modell",
|
|
|
|
|
|
"Phantomcloak19/qwen3-4b-dpo",
|
|
|
|
|
|
"A7med-Ame3/qwen3_merged_model",
|
|
|
|
|
|
"izzatiroza/qwen2.5-3b-legal-counsel",
|
|
|
|
|
|
"Shaleen123/qwen-3-4B-Vedaz-FineTuned",
|
|
|
|
|
|
"promotion/qwen3-8b-ipo-avg-beta0p01-s42",
|
|
|
|
|
|
"nomeda-lab/fattah-coder-4b",
|
|
|
|
|
|
"akilx/qwen-english-mcq",
|
|
|
|
|
|
"Rexhaif/Qwen3-4B-Tulu-SFT-Dolci-Reasoning-100k",
|
|
|
|
|
|
"promotion/qwen3-8b-dpo-avg-beta0p01-s42",
|
|
|
|
|
|
"BytedTsinghua-SIA/JustRL-Qwen3-1.7B",
|
|
|
|
|
|
"BytedTsinghua-SIA/JustRL-Qwen3-4B",
|
|
|
|
|
|
"e12ex2/Qwen3-1.7B-SigmaRL",
|
|
|
|
|
|
"viamr-project/qwen3-1.7b-amr-20260705-0708",
|
|
|
|
|
|
"t2ance/CodeRM-SFT-Warmup-Selection-1.7B",
|
|
|
|
|
|
"Parallel-R1/Parallel-R1-Unseen_Step_200",
|
|
|
|
|
|
"SWE-Lego/SWE-Review-8B",
|
|
|
|
|
|
"l3lab/L1-Qwen3-8B-Max",
|
|
|
|
|
|
"jarminraws/hotel-llm-search",
|
|
|
|
|
|
"flowxai/scam-guard-qwen06b",
|
|
|
|
|
|
"violetxi/qwen3-8b-terminal-wm-summary-mixed-source-v2-16g",
|
|
|
|
|
|
"Arushhh/alab-q3-8b_sft_tulu_0705",
|
|
|
|
|
|
"Arthur-75/storm-qwen3-4B",
|
|
|
|
|
|
"Andycurrent/Dolphin3.0-Llama3.1-8B",
|
|
|
|
|
|
"gauthierpiarrette/nl2jq-qwen3-0.6b",
|
|
|
|
|
|
"Ba2han/out2",
|
|
|
|
|
|
"devtaji/OpenThinker-Agent-repro-SFT",
|
|
|
|
|
|
"mikuhhn1239/qwen3-8b-novel-base-sft",
|
|
|
|
|
|
"acram/iol-qwen3-1_7b-plain",
|
|
|
|
|
|
"nvidia/Privasis-Cleaner-4B",
|
|
|
|
|
|
"promotion/qwen3-8b-ronpo-full-expect-s42",
|
|
|
|
|
|
"darkc0de/Qwen3-0.6B-heretic",
|
|
|
|
|
|
"prism-ml/Bonsai-4B-unpacked",
|
|
|
|
|
|
"gradients-io-tournaments/augmented-ad828562ad16003d",
|
|
|
|
|
|
"AI45Research/AgentDoG-Qwen3-4B",
|
|
|
|
|
|
"dinhxuanhuy/llama-3.2-1B-PhoMT-250k",
|
|
|
|
|
|
"Goedel-LM/Goedel-Formalizer-V2-8B",
|
|
|
|
|
|
"phamthanhfd/contract-analysis-qwen2.5-3b",
|
|
|
|
|
|
"font-info/qwen3-4b-sft-SGLang-RL",
|
|
|
|
|
|
"UnicomAI/Unichat-llama3.2-Chinese-1B",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-en-only",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-hi-only",
|
|
|
|
|
|
"DhruvalLabs/qwen3-8b-claude-agentic-fable5",
|
|
|
|
|
|
"yamatazen/Qwen3-HereticLM-4B",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-es-only",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-ko-only",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-ms-only",
|
|
|
|
|
|
"AttentioResearch/tally-8b-flagship",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-zh-only",
|
|
|
|
|
|
"922-CA/llama-2-7b-monika-v0.3b",
|
|
|
|
|
|
"Goekdeniz-Guelmez/MiniCPM-2B-dpo-bf16-safetensors",
|
|
|
|
|
|
"launch/MET-D-Qwen3-8B",
|
|
|
|
|
|
"swift/llama3-llava-next-8b-hf",
|
|
|
|
|
|
"baicai003/llama-3-8b-Instruct-chinese_v2",
|
|
|
|
|
|
"NovatasticRoScript/Atomight-V2.5-1.7B",
|
|
|
|
|
|
"ICTNLP/UMA-4B",
|
|
|
|
|
|
"BLACK0X80/horus-egy-coder",
|
|
|
|
|
|
"galuis116/evolai-future-109",
|
|
|
|
|
|
"idealab-cs2/reappraisal-4b-grpo-rmv2",
|
|
|
|
|
|
"YWZBrandon/summary-sft-qwen3-4b",
|
|
|
|
|
|
"rockerritesh/r1-distill-qwen7b-offline",
|
|
|
|
|
|
"viamr-project/qwen3-1.7b-amr-20260704-0113",
|
|
|
|
|
|
"yapeichang/Qwen2.5-7B-RM8B",
|
|
|
|
|
|
"yapeichang/Llama-3.1-8B-BLEUBERI",
|
|
|
|
|
|
"yapeichang/Llama-3.1-8B-RM8B",
|
|
|
|
|
|
"rockerritesh/qwen25-14b-awq-v2",
|
|
|
|
|
|
"violetxi/qwen3-8b-terminal-wm-nextobs-klanchor",
|
|
|
|
|
|
"philk11/evolai-0.4b",
|
|
|
|
|
|
"prism-ml/Ternary-Bonsai-8B-unpacked",
|
|
|
|
|
|
"NiuTrans/LMT-60-4B",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-e",
|
|
|
|
|
|
"leoeo999/AI-Legal-Chatbot",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-c",
|
|
|
|
|
|
"TejasviniC/IOL_V0",
|
|
|
|
|
|
"spitfire4794/Zupra-1.7-50M-Instruct-Ultra-Math-exp",
|
|
|
|
|
|
"Akkachai/Qwen3-0.6B-Base-CPT-Math",
|
|
|
|
|
|
"state-spaces/mamba-2.8b-hf",
|
|
|
|
|
|
"andyx10/Qwen2.5-1.5B-Instruct-NLA-L18-ar",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-0.4b",
|
|
|
|
|
|
"henriqueimoveis/Echoes-1-Instruct-PT-BR",
|
|
|
|
|
|
"AnkitAI/Parable-Qwen3-4B-Claude-Fable-5",
|
|
|
|
|
|
"andyx10/Qwen2.5-1.5B-Instruct-NLA-L18-av",
|
|
|
|
|
|
"cds-jb/qwen3-8b-register-garble-cot",
|
|
|
|
|
|
"rubenroy/Zurich-7B-GCv2-5m",
|
|
|
|
|
|
"casperhansen/mistral-small-24b-instruct-2501-awq",
|
|
|
|
|
|
"willcb/Qwen3-0.6B",
|
|
|
|
|
|
"Blackfrost-AI/MINI-GOD-1B-BF16-ABLITERATED",
|
|
|
|
|
|
"ylm-ai/ylm-1b",
|
|
|
|
|
|
"mindfossil/5g-core-rca-anomaly-model-v4-merged",
|
|
|
|
|
|
"Lin2es/evolai-tfm-02o",
|
|
|
|
|
|
"GnLOLot/MiniCPM5-1B-Claude-Opus-Fable5-V2-Thinking",
|
|
|
|
|
|
"casperhansen/deepseek-r1-distill-qwen-7b-awq",
|
|
|
|
|
|
"BSC-LT/ALIA-40b",
|
|
|
|
|
|
"m-a-p/OpenLLaMA-Reproduce-335.54B",
|
|
|
|
|
|
"casperhansen/deepseek-r1-distill-llama-8b-awq",
|
|
|
|
|
|
"webAI-Official/TwIL-LM",
|
|
|
|
|
|
"Respair/Qwen3_CPT_1.7B",
|
|
|
|
|
|
"facebook/opt-30b",
|
|
|
|
|
|
"OpenLLM-Ro/RoLlama3-8b-Instruct-2024-06-28",
|
|
|
|
|
|
"ai9stars/G9v3-3B",
|
|
|
|
|
|
"openbmb/MiniCPM5-1B",
|
|
|
|
|
|
"enochlev/MiniCPM-duplex",
|
|
|
|
|
|
"openbmb/AgentCPM-Report",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-4single-refusal",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-4single-hate",
|
|
|
|
|
|
"bibocat/qwen3-ner-grpo-v2-merged",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-4pair-refusal",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-4pair-hate",
|
|
|
|
|
|
"goldfish-models/pes_arab_100mb",
|
|
|
|
|
|
"pranjalthakz/physics-tutor-merged",
|
|
|
|
|
|
"jevonmao/llama31-8b-poker-mix-v1-step10k",
|
|
|
|
|
|
"vysri/SmolLM135M-IT-ConvFill",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-med-lora-null-v1-target",
|
|
|
|
|
|
"januschoy/druckenmiller-1.5b-v2",
|
|
|
|
|
|
"sirunchained/text-to-sql-model-v2",
|
|
|
|
|
|
"justasamthing/qwen2.5-3b-chat-alpaca-indonesian",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-med-lora-null-v2-oasst1",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-code-corda-oasst1",
|
|
|
|
|
|
"saital/iol-ai-2026-baseline",
|
|
|
|
|
|
"rahelrj/legal-chatbot-qlora-id",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-code-lora-null-v2-oasst1",
|
|
|
|
|
|
"NyayaLabs98/nyaya-3b-v3",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-code-lora-null-v1-target",
|
|
|
|
|
|
"ishikauniphore/student_SelectedGT_qwen7bins_nemotron_stem",
|
|
|
|
|
|
"ConnorYU/qwen3-8b-insecure-v6-verIH-local",
|
|
|
|
|
|
"DanielTobi0/iol-ai-2026",
|
|
|
|
|
|
"idoo0/qwen2.5-7b-legal-chatbot-sft-idoft",
|
|
|
|
|
|
"amank-root/demo-ddi-1.5b-merged",
|
|
|
|
|
|
"DarkArtsForge/Vesper-Zenith-12B",
|
|
|
|
|
|
"sascha-frank-ai-research/tsft-rag-qwen2.5-0.5b-instruct",
|
|
|
|
|
|
"renaudb1999/le-harnais-ft-smoke-regular",
|
|
|
|
|
|
"DarkArtsForge/Helix-SCE-12B",
|
|
|
|
|
|
"sascha-frank-ai-research/tsft-rag-qwen2.5-7b-instruct",
|
|
|
|
|
|
"sascha-frank-ai-research/tsft-rag-qwen2.5-1.5b-instruct",
|
|
|
|
|
|
"claye123/llama-2-13B",
|
|
|
|
|
|
"openbmb/MiniCPM4-0.5B",
|
|
|
|
|
|
"openbmb/MiniCPM4.1-8B",
|
|
|
|
|
|
"openbmb/MiniCPM4-MCP",
|
|
|
|
|
|
"openbmb/MiniCPM4-8B",
|
2026-07-29 14:26:16 +08:00
|
|
|
|
]
|
2026-07-23 14:27:36 +08:00
|
|
|
|
|
2026-07-29 14:26:16 +08:00
|
|
|
|
METAX_MODELS = [
|
2026-07-29 17:36:02 +08:00
|
|
|
|
"Xorbits/CodeLlama-7B-fp16",
|
|
|
|
|
|
"sstoica12/acquisition_student_llama-3_1-8b_bins_medmcqa_confidence",
|
|
|
|
|
|
"anime-sh/llama-3_1-8b-undial-bm25-10b-rebuttal",
|
|
|
|
|
|
"llm-jp/llm-jp-3-8x1.8b-instruct3",
|
|
|
|
|
|
"dphn/Dolphin3.0-R1-Mistral-24B",
|
|
|
|
|
|
"SicariusSicariiStuff/TinyLLama_0.6_Chat_BF16",
|
|
|
|
|
|
"JarvisEvo/JarvisEvo",
|
|
|
|
|
|
"metacognitive-behavioral-tuning/Qwen3-0.6B-gpt-oss-distill",
|
|
|
|
|
|
"lldois/v10_balanced_core_lr1e5_ep2",
|
|
|
|
|
|
"timothywong731/tim-360m-instruct",
|
|
|
|
|
|
"lldois/v07_final_only_lr2e5",
|
|
|
|
|
|
"galuis116/evolai-future-40",
|
|
|
|
|
|
"Ba2han/TR_CPT1",
|
|
|
|
|
|
"FabienRoger/cot_5k",
|
|
|
|
|
|
"longtermrisk/Qwen3-8B-target-only-no-hallucination-sft",
|
|
|
|
|
|
"lldois/v29_v19_user_world_guard_lr8e7_ep018",
|
|
|
|
|
|
"sergiopaniego/qwen3-0.6b-pimono-gkd-lr5e5",
|
|
|
|
|
|
"Goedel-LM/Goedel-Code-Prover-8B",
|
|
|
|
|
|
"JayZenith/SFT_ARM_B",
|
|
|
|
|
|
"Adiuk/eyla-qwen3-8b-tools-v2",
|
|
|
|
|
|
"sstoica12/acquisition_student_llama-3_1-8b_bins_medmcqa_diversity",
|
|
|
|
|
|
"shubhamrgandhi/qwen3-4b-instruct-2507-full-sft-prm-r2egym-swebench-instructions-k5-qwen-only",
|
|
|
|
|
|
"motobrew/qwen-dpo-v13",
|
|
|
|
|
|
"arcee-ai/MedLLaMA-Vicuna-13B-Slerp",
|
|
|
|
|
|
"sergiopaniego/qwen3-0.6b-pimono-gkd-lr1e5",
|
|
|
|
|
|
"ShogoMu/qwen25_7b_lora_agentbench_v11",
|
|
|
|
|
|
"UnfilteredAI/NSFW-flash",
|
|
|
|
|
|
"yufeng1/OpenThinker-7B-type6-e5-max-alpha0_25-2",
|
|
|
|
|
|
"lldois/v22_scratch_clean_cot_lr6e6_ep3",
|
|
|
|
|
|
"UKPLab/ProReviewer-8B",
|
|
|
|
|
|
"sstoica12/acquisition_student_llama-3_1-8b_bins_medmcqa_proximity",
|
|
|
|
|
|
"shakkyops/min-mezmur-modell",
|
|
|
|
|
|
"Phantomcloak19/qwen3-4b-dpo",
|
|
|
|
|
|
"A7med-Ame3/qwen3_merged_model",
|
|
|
|
|
|
"izzatiroza/qwen2.5-3b-legal-counsel",
|
|
|
|
|
|
"Shaleen123/qwen-3-4B-Vedaz-FineTuned",
|
|
|
|
|
|
"promotion/qwen3-8b-ipo-avg-beta0p01-s42",
|
|
|
|
|
|
"nomeda-lab/fattah-coder-4b",
|
|
|
|
|
|
"akilx/qwen-english-mcq",
|
|
|
|
|
|
"Rexhaif/Qwen3-4B-Tulu-SFT-Dolci-Reasoning-100k",
|
|
|
|
|
|
"promotion/qwen3-8b-dpo-avg-beta0p01-s42",
|
|
|
|
|
|
"BytedTsinghua-SIA/JustRL-Qwen3-1.7B",
|
|
|
|
|
|
"BytedTsinghua-SIA/JustRL-Qwen3-4B",
|
|
|
|
|
|
"e12ex2/Qwen3-1.7B-SigmaRL",
|
|
|
|
|
|
"viamr-project/qwen3-1.7b-amr-20260705-0708",
|
|
|
|
|
|
"t2ance/CodeRM-SFT-Warmup-Selection-1.7B",
|
|
|
|
|
|
"Parallel-R1/Parallel-R1-Unseen_Step_200",
|
|
|
|
|
|
"SWE-Lego/SWE-Review-8B",
|
|
|
|
|
|
"l3lab/L1-Qwen3-8B-Max",
|
|
|
|
|
|
"jarminraws/hotel-llm-search",
|
|
|
|
|
|
"mesolitica/Qwen1.5-0.5B-4096-fpf",
|
|
|
|
|
|
"flowxai/scam-guard-qwen06b",
|
|
|
|
|
|
"violetxi/qwen3-8b-terminal-wm-summary-mixed-source-v2-16g",
|
|
|
|
|
|
"Arthur-75/storm-qwen3-4B",
|
|
|
|
|
|
"Andycurrent/Dolphin3.0-Llama3.1-8B",
|
|
|
|
|
|
"gauthierpiarrette/nl2jq-qwen3-0.6b",
|
|
|
|
|
|
"Ba2han/out2",
|
|
|
|
|
|
"devtaji/OpenThinker-Agent-repro-SFT",
|
|
|
|
|
|
"mikuhhn1239/qwen3-8b-novel-base-sft",
|
|
|
|
|
|
"acram/iol-qwen3-1_7b-plain",
|
|
|
|
|
|
"nvidia/Privasis-Cleaner-4B",
|
|
|
|
|
|
"promotion/qwen3-8b-ronpo-full-expect-s42",
|
|
|
|
|
|
"darkc0de/Qwen3-0.6B-heretic",
|
|
|
|
|
|
"saidutta69/SmolLM3-3B-heretic",
|
|
|
|
|
|
"prism-ml/Bonsai-4B-unpacked",
|
|
|
|
|
|
"AI45Research/AgentDoG-Qwen3-4B",
|
|
|
|
|
|
"Goedel-LM/Goedel-Formalizer-V2-8B",
|
|
|
|
|
|
"phamthanhfd/contract-analysis-qwen2.5-3b",
|
|
|
|
|
|
"font-info/qwen3-4b-sft-SGLang-RL",
|
|
|
|
|
|
"UnicomAI/Unichat-llama3.2-Chinese-1B",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-en-only",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-hi-only",
|
|
|
|
|
|
"DhruvalLabs/qwen3-8b-claude-agentic-fable5",
|
|
|
|
|
|
"yamatazen/Qwen3-HereticLM-4B",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-es-only",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-ko-only",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-ms-only",
|
|
|
|
|
|
"AttentioResearch/tally-8b-flagship",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-zh-only",
|
|
|
|
|
|
"922-CA/llama-2-7b-monika-v0.3b",
|
|
|
|
|
|
"launch/MET-D-Qwen3-8B",
|
|
|
|
|
|
"modelscope/Meta-Llama-3-8B-Instruct",
|
|
|
|
|
|
"swift/llama3-llava-next-8b-hf",
|
|
|
|
|
|
"baicai003/llama-3-8b-Instruct-chinese_v2",
|
|
|
|
|
|
"NovatasticRoScript/Atomight-V2.5-1.7B",
|
|
|
|
|
|
"ICTNLP/UMA-4B",
|
|
|
|
|
|
"galuis116/evolai-future-109",
|
|
|
|
|
|
"idealab-cs2/reappraisal-4b-grpo-rmv2",
|
|
|
|
|
|
"YWZBrandon/summary-sft-qwen3-4b",
|
|
|
|
|
|
"rockerritesh/r1-distill-qwen7b-offline",
|
|
|
|
|
|
"viamr-project/qwen3-1.7b-amr-20260704-0113",
|
|
|
|
|
|
"yapeichang/Qwen2.5-7B-RM8B",
|
|
|
|
|
|
"yapeichang/Llama-3.1-8B-BLEUBERI",
|
|
|
|
|
|
"yapeichang/Llama-3.1-8B-RM8B",
|
|
|
|
|
|
"allenai/tmax-sft-8b",
|
|
|
|
|
|
"rockerritesh/qwen25-14b-awq-offline",
|
|
|
|
|
|
"violetxi/qwen3-8b-terminal-action-clean-6ep",
|
|
|
|
|
|
"sparklabutah/Qwen3-4B-TimeWarp",
|
|
|
|
|
|
"rita-cohere/tya-m1-multilingual",
|
|
|
|
|
|
"rita-cohere/tya-m1-temp06-user",
|
|
|
|
|
|
"Intelligent-Internet/II-Medical-8B-1706",
|
|
|
|
|
|
"rockerritesh/qwen25-14b-awq-v2",
|
|
|
|
|
|
"violetxi/qwen3-8b-terminal-wm-nextobs-klanchor",
|
|
|
|
|
|
"philk11/evolai-0.4b",
|
|
|
|
|
|
"prism-ml/Ternary-Bonsai-8B-unpacked",
|
|
|
|
|
|
"NiuTrans/LMT-60-4B",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-e",
|
|
|
|
|
|
"leoeo999/AI-Legal-Chatbot",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-b",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-c",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-d",
|
|
|
|
|
|
"Lin2es/evolai-tfm-04o",
|
|
|
|
|
|
"TejasviniC/IOL_V0",
|
|
|
|
|
|
"spitfire4794/Zupra-1.7-50M-Instruct-Ultra-Math-exp",
|
|
|
|
|
|
"andyx10/Qwen2.5-1.5B-Instruct-NLA-L18-ar",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-0.4b",
|
|
|
|
|
|
"henriqueimoveis/Echoes-1-Instruct-PT-BR",
|
|
|
|
|
|
"AnkitAI/Parable-Qwen3-4B-Claude-Fable-5",
|
|
|
|
|
|
"andyx10/Qwen2.5-1.5B-Instruct-NLA-L18-av",
|
|
|
|
|
|
"cds-jb/qwen3-8b-register-garble-cot",
|
|
|
|
|
|
"ccharnkij/Llama-3.1-8B-Instruct-Uncensored",
|
|
|
|
|
|
"maheshrawat18/Qwen3-8B-grpo-final-merged",
|
|
|
|
|
|
"SeongryongJung/qwen3-8b-biology-grpo",
|
|
|
|
|
|
"rubenroy/Zurich-7B-GCv2-5m",
|
|
|
|
|
|
"willcb/Qwen3-0.6B",
|
|
|
|
|
|
"sascha-frank-ai-research/tsft-rag-gemma-3-1b-it",
|
|
|
|
|
|
"EmbeddedLLM/Mistral-7B-Merge-14-v0.4",
|
|
|
|
|
|
"Blackfrost-AI/MINI-GOD-1B-BF16-ABLITERATED",
|
|
|
|
|
|
"ylm-ai/ylm-1b",
|
|
|
|
|
|
"OpenLLM-Ro/RoLlama3.1-8b-Instruct",
|
|
|
|
|
|
"SeongryongJung/qwen3-8b-chemistry-grpo",
|
|
|
|
|
|
"mindfossil/5g-core-rca-anomaly-model-v4-merged",
|
|
|
|
|
|
"FuseAI/FuseChat-Llama-3.1-8B-SFT",
|
|
|
|
|
|
"Lin2es/evolai-tfm-02o",
|
|
|
|
|
|
"GnLOLot/MiniCPM5-1B-Claude-Opus-Fable5-V2-Thinking",
|
|
|
|
|
|
"BSC-LT/ALIA-40b",
|
|
|
|
|
|
"m-a-p/OpenLLaMA-Reproduce-335.54B",
|
|
|
|
|
|
"webAI-Official/TwIL-LM",
|
|
|
|
|
|
"aifeifei798/llama3-8B-DarkIdol-2.2-Uncensored-1048K",
|
|
|
|
|
|
"aifeifei798/llama3-8B-DarkIdol-2.1-Uncensored-1048K",
|
|
|
|
|
|
"Respair/Qwen3_CPT_1.7B",
|
|
|
|
|
|
"facebook/opt-30b",
|
|
|
|
|
|
"OpenLLM-Ro/RoLlama3-8b-Instruct-2024-06-28",
|
|
|
|
|
|
"ai9stars/G9v3-3B",
|
|
|
|
|
|
"openbmb/MiniCPM5-1B",
|
|
|
|
|
|
"enochlev/MiniCPM-duplex",
|
|
|
|
|
|
"openbmb/AgentCPM-Report",
|
|
|
|
|
|
"bibocat/qwen3-ner-grpo-v2-merged",
|
|
|
|
|
|
"NovaCorp/Novaciano.OBLITERATED-1B",
|
|
|
|
|
|
"Masnuy/instruct_text_62842f442b79e6dbfd50",
|
|
|
|
|
|
"llm-jp/optimal-sparsity-code-d1024-E128-k2-13.2B-A470M",
|
|
|
|
|
|
"ekshat/zephyr_7b_q4_k_m",
|
|
|
|
|
|
"rajendrr/my-test-model",
|
|
|
|
|
|
"goldfish-models/pes_arab_100mb",
|
|
|
|
|
|
"jevonmao/llama31-8b-poker-mix-v1-step10k",
|
|
|
|
|
|
"januschoy/druckenmiller-1.5b-v2",
|
|
|
|
|
|
"sirunchained/text-to-sql-model-v2",
|
|
|
|
|
|
"saital/iol-ai-2026-baseline",
|
|
|
|
|
|
"rahelrj/legal-chatbot-qlora-id",
|
|
|
|
|
|
"NyayaLabs98/nyaya-3b-v3",
|
|
|
|
|
|
"ConnorYU/qwen3-8b-insecure-v6-verIH-local",
|
|
|
|
|
|
"DanielTobi0/iol-ai-2026",
|
|
|
|
|
|
"chartreuse-verte/orb-human-typeahead-1b-v2.1",
|
|
|
|
|
|
"BigRatz/LOL-AI-2026-V2",
|
|
|
|
|
|
"yaqi2/Qwen3-1.7B-ref",
|
|
|
|
|
|
"chartreuse-verte/orb-human-typeahead-350m-v1.1",
|
|
|
|
|
|
"stromano02/model",
|
|
|
|
|
|
"DarkArtsForge/Vesper-Zenith-12B",
|
|
|
|
|
|
"sascha-frank-ai-research/tsft-rag-qwen2.5-0.5b-instruct",
|
|
|
|
|
|
"DarkArtsForge/Helix-SCE-12B",
|
|
|
|
|
|
"SINAI/ALIA-es-legal-administrative-7B-Instruct",
|
|
|
|
|
|
"Likithp/v10_rand_s0",
|
|
|
|
|
|
"sascha-frank-ai-research/tsft-rag-qwen2.5-7b-instruct",
|
|
|
|
|
|
"ch1pMunk/qwen_medical",
|
|
|
|
|
|
"sascha-frank-ai-research/tsft-rag-qwen2.5-1.5b-instruct",
|
|
|
|
|
|
"claye123/llama-2-13B",
|
2026-06-10 21:42:41 +08:00
|
|
|
|
]
|
|
|
|
|
|
|
2026-07-29 15:16:34 +08:00
|
|
|
|
KUNLUNXIN_MODELS = [
|
2026-07-29 17:36:02 +08:00
|
|
|
|
"lldois/v10_balanced_core_lr1e5_ep2",
|
|
|
|
|
|
"timothywong731/tim-360m-instruct",
|
|
|
|
|
|
"lldois/v07_final_only_lr2e5",
|
|
|
|
|
|
"galuis116/evolai-future-40",
|
|
|
|
|
|
"longtermrisk/Qwen3-8B-target-only-no-hallucination-sft",
|
|
|
|
|
|
"lldois/v29_v19_user_world_guard_lr8e7_ep018",
|
|
|
|
|
|
"sergiopaniego/qwen3-0.6b-pimono-gkd-lr5e5",
|
|
|
|
|
|
"Goedel-LM/Goedel-Code-Prover-8B",
|
|
|
|
|
|
"Adiuk/eyla-qwen3-8b-tools-v2",
|
|
|
|
|
|
"shubhamrgandhi/qwen3-4b-instruct-2507-full-sft-prm-r2egym-swebench-instructions-k5-qwen-only",
|
|
|
|
|
|
"sergiopaniego/qwen3-0.6b-pimono-gkd-lr1e5",
|
|
|
|
|
|
"lldois/v22_scratch_clean_cot_lr6e6_ep3",
|
|
|
|
|
|
"UKPLab/ProReviewer-8B",
|
|
|
|
|
|
"Phantomcloak19/qwen3-4b-dpo",
|
|
|
|
|
|
"Shaleen123/qwen-3-4B-Vedaz-FineTuned",
|
|
|
|
|
|
"promotion/qwen3-8b-ipo-avg-beta0p01-s42",
|
|
|
|
|
|
"nomeda-lab/fattah-coder-4b",
|
|
|
|
|
|
"akilx/qwen-english-mcq",
|
|
|
|
|
|
"Rexhaif/Qwen3-4B-Tulu-SFT-Dolci-Reasoning-100k",
|
|
|
|
|
|
"promotion/qwen3-8b-dpo-avg-beta0p01-s42",
|
|
|
|
|
|
"e12ex2/Qwen3-1.7B-SigmaRL",
|
|
|
|
|
|
"viamr-project/qwen3-1.7b-amr-20260705-0708",
|
|
|
|
|
|
"t2ance/CodeRM-SFT-Warmup-Selection-1.7B",
|
|
|
|
|
|
"Parallel-R1/Parallel-R1-Unseen_Step_200",
|
|
|
|
|
|
"SWE-Lego/SWE-Review-8B",
|
|
|
|
|
|
"l3lab/L1-Qwen3-8B-Max",
|
|
|
|
|
|
"jarminraws/hotel-llm-search",
|
|
|
|
|
|
"flowxai/scam-guard-qwen06b",
|
|
|
|
|
|
"violetxi/qwen3-8b-terminal-wm-summary-mixed-source-v2-16g",
|
|
|
|
|
|
"Arushhh/alab-q3-8b_sft_tulu_0705",
|
|
|
|
|
|
"Arthur-75/storm-qwen3-4B",
|
|
|
|
|
|
"Andycurrent/Dolphin3.0-Llama3.1-8B",
|
|
|
|
|
|
"gauthierpiarrette/nl2jq-qwen3-0.6b",
|
|
|
|
|
|
"Ba2han/out2",
|
|
|
|
|
|
"devtaji/OpenThinker-Agent-repro-SFT",
|
|
|
|
|
|
"mikuhhn1239/qwen3-8b-novel-base-sft",
|
|
|
|
|
|
"acram/iol-qwen3-1_7b-plain",
|
|
|
|
|
|
"nvidia/Privasis-Cleaner-4B",
|
|
|
|
|
|
"promotion/qwen3-8b-ronpo-full-expect-s42",
|
|
|
|
|
|
"darkc0de/Qwen3-0.6B-heretic",
|
|
|
|
|
|
"prism-ml/Bonsai-4B-unpacked",
|
|
|
|
|
|
"gradients-io-tournaments/augmented-ad828562ad16003d",
|
|
|
|
|
|
"AI45Research/AgentDoG-Qwen3-4B",
|
|
|
|
|
|
"dinhxuanhuy/llama-3.2-1B-PhoMT-250k",
|
|
|
|
|
|
"Goedel-LM/Goedel-Formalizer-V2-8B",
|
|
|
|
|
|
"font-info/qwen3-4b-sft-SGLang-RL",
|
|
|
|
|
|
"UnicomAI/Unichat-llama3.2-Chinese-1B",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-en-only",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-hi-only",
|
|
|
|
|
|
"DhruvalLabs/qwen3-8b-claude-agentic-fable5",
|
|
|
|
|
|
"yamatazen/Qwen3-HereticLM-4B",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-es-only",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-ko-only",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-ms-only",
|
|
|
|
|
|
"launch/MET-D-Qwen3-4B-zh-only",
|
|
|
|
|
|
"922-CA/llama-2-7b-monika-v0.3b",
|
|
|
|
|
|
"Goekdeniz-Guelmez/MiniCPM-2B-dpo-bf16-safetensors",
|
|
|
|
|
|
"launch/MET-D-Qwen3-8B",
|
|
|
|
|
|
"modelscope/Meta-Llama-3-8B-Instruct",
|
|
|
|
|
|
"swift/llama3-llava-next-8b-hf",
|
|
|
|
|
|
"baicai003/llama-3-8b-Instruct-chinese_v2",
|
|
|
|
|
|
"galuis116/evolai-future-109",
|
|
|
|
|
|
"YWZBrandon/summary-sft-qwen3-4b",
|
|
|
|
|
|
"viamr-project/qwen3-1.7b-amr-20260704-0113",
|
|
|
|
|
|
"yapeichang/Qwen2.5-7B-RM8B",
|
|
|
|
|
|
"yapeichang/Llama-3.1-8B-BLEUBERI",
|
|
|
|
|
|
"yapeichang/Llama-3.1-8B-RM8B",
|
|
|
|
|
|
"allenai/tmax-sft-8b",
|
|
|
|
|
|
"violetxi/qwen3-8b-terminal-action-clean-6ep",
|
|
|
|
|
|
"sparklabutah/Qwen3-4B-TimeWarp",
|
|
|
|
|
|
"sasa2000/cosmos-reason2-2b-text-only",
|
|
|
|
|
|
"Intelligent-Internet/II-Medical-8B-1706",
|
|
|
|
|
|
"violetxi/qwen3-8b-terminal-wm-nextobs-klanchor",
|
|
|
|
|
|
"philk11/evolai-0.4b",
|
|
|
|
|
|
"prism-ml/Ternary-Bonsai-8B-unpacked",
|
|
|
|
|
|
"NiuTrans/LMT-60-4B",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-e",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-b",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-c",
|
|
|
|
|
|
"icedsoylatte/qwen25-3b-chai-roleplay-sft-v1",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-d",
|
|
|
|
|
|
"Lin2es/evolai-tfm-04o",
|
|
|
|
|
|
"Akkachai/Qwen3-0.6B-Base-CPT-Math",
|
|
|
|
|
|
"andrebarrosilva1123/evolai-0.4b",
|
|
|
|
|
|
"ccharnkij/Llama-3.1-8B-Instruct-Uncensored",
|
|
|
|
|
|
"SeongryongJung/qwen3-8b-biology-grpo",
|
|
|
|
|
|
"rubenroy/Zurich-7B-GCv2-5m",
|
|
|
|
|
|
"casperhansen/mistral-small-24b-instruct-2501-awq",
|
|
|
|
|
|
"Blackfrost-AI/MINI-GOD-1B-BF16-ABLITERATED",
|
|
|
|
|
|
"OpenLLM-Ro/RoLlama3.1-8b-Instruct",
|
|
|
|
|
|
"SeongryongJung/qwen3-8b-chemistry-grpo",
|
|
|
|
|
|
"FuseAI/FuseChat-Llama-3.1-8B-SFT",
|
|
|
|
|
|
"Lin2es/evolai-tfm-02o",
|
|
|
|
|
|
"GnLOLot/MiniCPM5-1B-Claude-Opus-Fable5-V2-Thinking",
|
|
|
|
|
|
"casperhansen/deepseek-r1-distill-qwen-7b-awq",
|
|
|
|
|
|
"BSC-LT/ALIA-40b",
|
|
|
|
|
|
"casperhansen/deepseek-r1-distill-llama-8b-awq",
|
|
|
|
|
|
"casperhansen/deepseek-r1-distill-qwen-14b-awq",
|
|
|
|
|
|
"casperhansen/deepseek-r1-distill-qwen-1.5b-awq",
|
|
|
|
|
|
"aifeifei798/llama3-8B-DarkIdol-2.2-Uncensored-1048K",
|
|
|
|
|
|
"Respair/Qwen3_CPT_1.7B",
|
|
|
|
|
|
"OpenLLM-Ro/RoLlama3-8b-Instruct-2024-06-28",
|
|
|
|
|
|
"openbmb/MiniCPM5-1B",
|
|
|
|
|
|
"enochlev/MiniCPM-duplex",
|
|
|
|
|
|
"openbmb/AgentCPM-Report",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-4single-refusal",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-4single-hate",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-4pair-refusal",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-4pair-hate",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-2single-refusal",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-2single-hate",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-2pair-refusal",
|
|
|
|
|
|
"thoughtworks/backdoor-gemma2-2b-2pair-hate",
|
|
|
|
|
|
"ekshat/zephyr_7b_q4_k_m",
|
|
|
|
|
|
"goldfish-models/pes_arab_100mb",
|
|
|
|
|
|
"pranjalthakz/physics-tutor-merged",
|
|
|
|
|
|
"jevonmao/llama31-8b-poker-mix-v1-step10k",
|
|
|
|
|
|
"vysri/SmolLM135M-IT-ConvFill",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-med-lora-null-v1-target",
|
|
|
|
|
|
"justasamthing/qwen2.5-3b-chat-alpaca-indonesian",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-med-lora-null-v2-oasst1",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-code-corda-oasst1",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-code-lora-null-v2-oasst1",
|
|
|
|
|
|
"wz7475/llama-3.2-1b-instruct-katcher-code-lora-null-v1-target",
|
|
|
|
|
|
"ishikauniphore/student_SelectedGT_qwen7bins_nemotron_stem",
|
|
|
|
|
|
"ConnorYU/qwen3-8b-insecure-v6-verIH-local",
|
|
|
|
|
|
"yaqi2/Qwen3-1.7B-ref",
|
|
|
|
|
|
"affandymurad/legal-ft-grpo",
|
|
|
|
|
|
"idoo0/qwen2.5-7b-legal-chatbot-sft-idoft",
|
|
|
|
|
|
"amank-root/demo-ddi-1.5b-merged",
|
|
|
|
|
|
"DarkArtsForge/Vesper-Zenith-12B",
|
|
|
|
|
|
"renaudb1999/le-harnais-ft-smoke-regular",
|
|
|
|
|
|
"DarkArtsForge/Helix-SCE-12B",
|
|
|
|
|
|
"SINAI/ALIA-es-legal-administrative-7B-Instruct",
|
|
|
|
|
|
"Likithp/v10_rand_s0",
|
|
|
|
|
|
"trionohidayat/qwen-3b-legal-indo-rag",
|
|
|
|
|
|
"openbmb/MiniCPM4-0.5B",
|
|
|
|
|
|
"openbmb/MiniCPM4.1-8B",
|
|
|
|
|
|
"openbmb/MiniCPM4-MCP",
|
|
|
|
|
|
"openbmb/MiniCPM4-8B",
|
2026-07-29 15:16:34 +08:00
|
|
|
|
]
|
|
|
|
|
|
|
2026-08-04 20:54:10 +08:00
|
|
|
|
PPU_MODELS = [
|
2026-08-18 15:43:55 +08:00
|
|
|
|
"beberik/Nyxene-v3-11B",
|
|
|
|
|
|
"vclmax/nemo-12b-expansion-public-v1",
|
|
|
|
|
|
"Huysun29/cbt-gemma2-9b-v2",
|
|
|
|
|
|
"MaliDDD/openai_us4_500_8B",
|
|
|
|
|
|
"mychen76/mistral-7b-merged-dare_6x7",
|
|
|
|
|
|
"LLM-Research/tulu-v2.5-ppo-13b-hh-rlhf-60k",
|
|
|
|
|
|
"Undi95/LewdEngine",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-shuffle_dyck_steps250-seed208-preserve_emb",
|
|
|
|
|
|
"jeiku/32K_Selfbot",
|
|
|
|
|
|
"jordanpainter/diallm-qwen-gspo-aus",
|
|
|
|
|
|
"famepram/qwen2.5-1b",
|
|
|
|
|
|
"netcat420/MFANN3b",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-music_steps100-seed208-preserve_emb",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto_w1_mmd-w1lam8.4e-4_mmdrho8.4e-4_kr0.1-seed3",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto-nolam-seed3",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto-nolam-seed0",
|
|
|
|
|
|
"OpenAssistant/pythia-12b-pre-v8-12.5k-steps",
|
|
|
|
|
|
"notcvnt/Qwen3-4B-Thinking-2507-heretic",
|
|
|
|
|
|
"sequelbox/Qwen3-4B-PlumEsper",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-shuffle_dyck_steps250-seed324-preserve_emb",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto_w1-w1lam9.68e-4-seed3",
|
|
|
|
|
|
"build-small-hackathon/deal_sft_4B_hard",
|
|
|
|
|
|
"Undi95/CodeEngine",
|
|
|
|
|
|
"Undi95/Emerald-13B",
|
|
|
|
|
|
"BSC-LT/Checkpoint_2b_instructed_beta",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-music_steps500-seed1024-preserve_emb",
|
|
|
|
|
|
"gradients-io-tournaments/augmented-b8fb794abce85014",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto_kl-klam0.0333_beta0.1-seed3",
|
|
|
|
|
|
"MayaPH/FinOPT-Franklin",
|
|
|
|
|
|
"UnicomAI/Unichat-DeepSeek-R1-Distill-Qwen-7B-Safe-bf16",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-nca_steps500-seed208-preserve_emb",
|
|
|
|
|
|
"RylanSchaeffer/mem_Qwen3-93M_minerva_math_rep_0_sbst_1.0000_epch_1_ot_1",
|
|
|
|
|
|
"fhai50032/Mistral-4B-FT-2",
|
|
|
|
|
|
"skumar9/Llama-medx_v3.1",
|
|
|
|
|
|
"Leoymoffical/Qwen3-1.7-Leoym-MusaBase",
|
|
|
|
|
|
"Torero96Dev/Cupid-Qwen3-4B-v0.2",
|
|
|
|
|
|
"openbmb/Eurus-7b-kto",
|
|
|
|
|
|
"andquant/code",
|
|
|
|
|
|
"LittleTReee/qwen3-1.7b-littletree",
|
|
|
|
|
|
"CodeStrux-Tech/tac-1",
|
|
|
|
|
|
"Yukang/Qwen2.5-3B-Open-R1-GRPO",
|
|
|
|
|
|
"RyotaroOKabe/ope_gpt2_v1.1",
|
|
|
|
|
|
"OpenBMB/BitCPM-CANN-0.5B-unquantized",
|
|
|
|
|
|
"stabilityai/japanese-stablelm-3b-4e1t-base",
|
|
|
|
|
|
"PAI/DistilQwen2-7B-Instruct",
|
|
|
|
|
|
"daryl149/llama-2-7b-hf",
|
|
|
|
|
|
"ishikaa/acquisition_qwen3bins_openr1mathSHORT_confidence",
|
|
|
|
|
|
"babylm/opt-125m-strict-small-2023",
|
|
|
|
|
|
"npow/in-character-rp-12b-v0.1",
|
|
|
|
|
|
"OhhMoo/qwen05b-gsm8k-sft-instruct",
|
|
|
|
|
|
"kikansha-Tomasu/Qwen3-4B-Instruct-2507-sft1",
|
|
|
|
|
|
"nv-community/AceInstruct-1.5B",
|
|
|
|
|
|
"suayptalha/FastLlama-3.2-1B-Instruct",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto_mmd-mmdrho5.19e-3_kr0.1-seed6",
|
|
|
|
|
|
"nm-testing/SparseLlama-2-7b-evol-code-alpaca-pruned_70",
|
|
|
|
|
|
"openeurollm/oellm-9b-128k-theta32m-v3",
|
|
|
|
|
|
"Josephgflowers/Tinyllama-Cinder-1.3B-Reason-Test.2",
|
|
|
|
|
|
"vasishta/FastContext-1.0-4B-SFT-Fork",
|
|
|
|
|
|
"nothingiisreal/MN-12B-Starcannon-v3",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto_w1-w1lam9.68e-4-seed6",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto_w1_mmd-w1lam8.4e-4_mmdrho8.4e-4_kr0.1-seed6",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto-nolam-seed6",
|
|
|
|
|
|
"abidlabs/gemma-3-270m-text2sql",
|
|
|
|
|
|
"allknowingroger/HomerSlerp4-7B",
|
|
|
|
|
|
"TMLR-Group-HF/GT-Qwen3-8B-Base-MATH",
|
|
|
|
|
|
"usail-hkust/JailJudge-guard",
|
|
|
|
|
|
"appvoid/palmer-002",
|
|
|
|
|
|
"aisingapore/Llama-SEA-LION-v2-8B",
|
|
|
|
|
|
"beaugogh/pythia-1.4b-deduped-sharegpt",
|
|
|
|
|
|
"amu870/test-v2.1-dpo",
|
|
|
|
|
|
"taharmasmaliyev07/Qwen-3-4B-spell-checker",
|
|
|
|
|
|
"SloppyModels/Llama3.2-1B-NoGQA-GPTNeoXTokenizer",
|
|
|
|
|
|
"sashaboguraev/pythia-1b-ppt-control_nca_steps1000_1b-seed208-preserve_emb",
|
|
|
|
|
|
"Novaciano/qp-3.2-1B",
|
|
|
|
|
|
"SvalTek/L3.1-TheGrimoire-8B-Base-Test0",
|
|
|
|
|
|
"bgbgbrt/Thought-Aligner-7B-v1.0",
|
|
|
|
|
|
"OpenBuddy/openbuddy-deepseekcoder-6b-v16.1-32k",
|
|
|
|
|
|
"AI-ModelScope/granite-7b-base",
|
|
|
|
|
|
"intervitens/mini-magnum-12b-v1.1",
|
|
|
|
|
|
"YudhaPalulun/qwen2.5-1.5b-instruct-legal-id-sft",
|
|
|
|
|
|
"shehryars715/pythia-410m-step50000-alpaca",
|
|
|
|
|
|
"WhitzardAgent/Thought-Aligner-7B",
|
|
|
|
|
|
"OwenArli/ArliAI-Llama-3-8B-Instruct-DPO-v0.2",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-nca_steps1000-seed208-preserve_emb",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto_w1_mmd-w1lam8.4e-4_mmdrho8.4e-4_kr0.1-seed5",
|
|
|
|
|
|
"Dahoas/pythia-1B-static-sft",
|
|
|
|
|
|
"Xorbits/vicuna-7b-v1.5",
|
|
|
|
|
|
"abocide/Qwen2.5-7B-Instruct-R1-forfinance",
|
|
|
|
|
|
"NewstaR/Starlight-13B",
|
|
|
|
|
|
"aifeifei799/Llama-3.1-8B-Instruct-Fei-v1-Uncensored",
|
|
|
|
|
|
"LorenaYannnnn/general_reward-Qwen3-0.6B-baseline_all_tokens-seed_1",
|
|
|
|
|
|
"BAAI/Infinity-Instruct-7M-Gen-mistral-7B",
|
|
|
|
|
|
"shade1301/pgabl-legal-rag-qwen25-1_5b-v1",
|
|
|
|
|
|
"Neiwawastaken/legal-chatbot-llama3B-sft-merged",
|
|
|
|
|
|
"EphAsad/Atem-8B",
|
|
|
|
|
|
"sashaboguraev/pythia-1b-ppt-random_numbers_steps500_1b-seed208-preserve_emb",
|
|
|
|
|
|
"cs-552-2026-aaty/general_knowledge_model",
|
|
|
|
|
|
"ccui46/cookingworld_per_chunk_act_q3_tokfix_diffPrompt_higherLR_tformerPin_2000",
|
|
|
|
|
|
"llm-jp/llm-jp-3-7.2b-instruct2",
|
|
|
|
|
|
"01ai/Yi-9B",
|
|
|
|
|
|
"shisa-ai/ablation-168-a163.dpo.finaldpo.if25-shisa-v2-llama-3.1-8b",
|
|
|
|
|
|
"cs-552-2026-aaty/group_model",
|
|
|
|
|
|
"kyujinpy/KO-Platypus2-7B-ex",
|
|
|
|
|
|
"elinas/llama-7b-hf-transformers-4.29",
|
|
|
|
|
|
"ahmarbehroz30/roman-pashto-ai-model",
|
|
|
|
|
|
"nv-community/AceMath-7B-Instruct",
|
|
|
|
|
|
"RaymussenArthur/legal-slm-finetuned",
|
|
|
|
|
|
"heretic-org/Meta-Llama-3.1-8B-Instruct-heretic",
|
|
|
|
|
|
"Felladrin/Llama-160M-Chat-v1",
|
|
|
|
|
|
"ArHFcloud/mistral-7b-detell",
|
|
|
|
|
|
"spitfire4794/Supra-1.5-50M-Base-exp-Distill",
|
|
|
|
|
|
"princeton-nlp/lm-1.3B-select_30B_tokens_by-educational_value-sample_with_temperature1.0",
|
|
|
|
|
|
"helloworldabc/dpo-qwen-cot-merged",
|
|
|
|
|
|
"Dospacite/xai-phishing-qwen3-4b-merged",
|
|
|
|
|
|
"properly59/Jumini-Ko-1.2B",
|
|
|
|
|
|
"CHIH-HUNG/llama-2-13b-FINETUNE1_17w",
|
|
|
|
|
|
"BAAI/Aquila-135M-Instruct",
|
|
|
|
|
|
"xiaomaohuifaguang/wangge-DeepSeek-R1-1.5B",
|
|
|
|
|
|
"openeurollm/oellm-9b-128k-theta32m-prelude",
|
|
|
|
|
|
"AI-ModelScope/Llama-3.2-1B-Instruct",
|
|
|
|
|
|
"beaugogh/Llama2-7b-openorca-mc-v1",
|
|
|
|
|
|
"yunjae-won/mpq3_llama8b_sft_dpo_beta1e-1_step3328",
|
|
|
|
|
|
"CHIH-HUNG/llama-2-13b-FINETUNE4_3.8w-r16-q_k_v_o_gate_up_down",
|
|
|
|
|
|
"BAAI/Aquila-135M",
|
|
|
|
|
|
"yunjae-won/mpq3_llama8b_sft_dpo_beta1e-1_step2816",
|
|
|
|
|
|
"ewald1976/Orionian-Dreams-Bar-and-Cafe-12B",
|
|
|
|
|
|
"lgaalves/mistral-7b_open_platypus",
|
|
|
|
|
|
"Valencio/LLM_course_eli5_clm-model",
|
|
|
|
|
|
"stefra/qwen_fm_2k",
|
|
|
|
|
|
"dhmeltzer/Llama-2-13b-hf-ds_eli5_1024_r_64_alpha_16_merged",
|
|
|
|
|
|
"CohereForAI/aya-expanse-8b",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-control_shuffle_dyck_steps500-seed1024-preserve_emb",
|
|
|
|
|
|
"ShadowFall09/FANNO",
|
|
|
|
|
|
"h2oai/h2ogpt-oig-oasst1-512-6_9b",
|
|
|
|
|
|
"Xorbits/opt-125m",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-control_shuffle_dyck_steps1000-seed324-preserve_emb",
|
|
|
|
|
|
"linjh1118/WidsoMenter-8B",
|
|
|
|
|
|
"Dogoo3/Aletheia-12B",
|
|
|
|
|
|
"Huysun29/cbt-qwen2.5-7b-v2",
|
|
|
|
|
|
"hammh0a/Hala-9B",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-control_shuffle_dyck_steps1000-seed208-preserve_emb",
|
|
|
|
|
|
"aryan-kolapkar/MathReasoner-Mini-1.5b",
|
|
|
|
|
|
"OpenBuddy/OpenBuddy-R10528DistillQwen-4B-v27.2",
|
|
|
|
|
|
"jefferylovely/SuperThetaMaven",
|
|
|
|
|
|
"Azzindani/Deepseek_ID_Legal_Preview",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-shuffle_dyck_steps250-seed1024",
|
|
|
|
|
|
"VGLee/qwen1.5-4b-universal_ner-galore",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-control_shuffle_dyck_steps500-seed208-preserve_emb",
|
|
|
|
|
|
"RedRenisa/PGABL-Renisa-Assyifa-Putri-legal-chatbot-grpo",
|
|
|
|
|
|
"abhinand/tamil-llama-13b-base-v0.1",
|
|
|
|
|
|
"jmichaelov/parc-pythia-seed0",
|
|
|
|
|
|
"mxcui/vanilla-imdb-ppo-prop0.2-alpha1.0-seed42-mean_kl0.1-EleutherAI-pythia-160m",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-music_steps100-seed1024-preserve_emb",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-music_steps500-seed208-preserve_emb",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-music_steps500-seed324-preserve_emb",
|
|
|
|
|
|
"DarkArtsForge/Magistaroth-24B-v1",
|
|
|
|
|
|
"jondurbin/airoboros-7b-gpt4-1.1",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-control_shuffle_dyck_steps500-seed324-preserve_emb",
|
|
|
|
|
|
"extraltodeus/llamachill_13b",
|
|
|
|
|
|
"CHIH-HUNG/llama-2-13b-FINETUNE4_3.8w-r16-gate_up_down",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-control_shuffle_dyck_steps1000-seed1024-preserve_emb",
|
|
|
|
|
|
"mxcui/maxmin-imdb-ppo-prop0.2-alpha1.0-seed42-mean_kl0.1-EleutherAI-pythia-160m",
|
|
|
|
|
|
"ResplendentAI/DaturaCookie_7B",
|
|
|
|
|
|
"gguk2on/olmo2-7b-rlar_g8_b384_math_0.20.08",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-control_nca_steps1000-seed208-preserve_emb",
|
|
|
|
|
|
"gradients-io-tournaments/tournament-tourn_c03a612f287687e0_20260713-f0783105-5d78-4968-9e88-d26d8bcca0c1-5EeLcV3L",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto_w1-w1lam9.68e-4-seed2",
|
|
|
|
|
|
"DavidAU/L3.1-RP-Hero-BigTalker-8B",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto_kl-klam0.0333_beta0.1-seed2",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto_mmd-mmdrho5.19e-3_kr0.1-seed2",
|
|
|
|
|
|
"saarvajanik/facebook-opt-6.7b-gqa-ub-16-best-for-KV-cache",
|
|
|
|
|
|
"ChaoticNeutrals/Prima-LelantaclesV7-experimental-7b",
|
|
|
|
|
|
"C10X/Nanbeige4-3B-Thinking-2511-Claude-4.5-Opus-High-Reasoning-Distill",
|
|
|
|
|
|
"AVeryRealHuman/DialoGPT-small-TonyStark",
|
|
|
|
|
|
"mxcui/vanilla-imdb-ppo-prop0.2-alpha1.0-seed42-kl0.15-EleutherAI-pythia-160m-lr7e-6",
|
|
|
|
|
|
"mncai/Mistral-7B-Instruct-v0.2-1st-NWS-KoOrca-5k-2nd-LaAdMoAl-o10u5",
|
|
|
|
|
|
"TheBloke/CodeLlama-7B-Python-fp16",
|
|
|
|
|
|
"cs-552-2026-MMRF/DARE",
|
|
|
|
|
|
"maywell/Synatra_TbST02M_IN01",
|
|
|
|
|
|
"transformers-community/dola",
|
|
|
|
|
|
"Secbone/llama-2-13B-instructed",
|
|
|
|
|
|
"cs-552-2026-MMRF/TIES",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-nca_steps100-seed324-preserve_emb",
|
|
|
|
|
|
"linsong8208/trainer_output",
|
|
|
|
|
|
"Gueule-d-ange/aup-fullft-kto_w1_mmd-w1lam8.4e-4_mmdrho8.4e-4_kr0.1-seed2",
|
|
|
|
|
|
"arcee-ai/Patent-Llama-7B-Chat-Slerp",
|
|
|
|
|
|
"aJupyter/EmoLLM_Qwen1_5-0_5B-Chat_full_sft",
|
|
|
|
|
|
"RefalMachine/ruadapt_qwen2.5_7B_ext_u48_instruct",
|
|
|
|
|
|
"m-a-p/OpenLLaMA-Reproduce-2030.04B",
|
|
|
|
|
|
"FlagRelease/Qwen2-7B-Instruct-FlagOS",
|
|
|
|
|
|
"chartreuse-verte/orb-human-typeahead-1b-v2",
|
|
|
|
|
|
"cs-552-2026-MMRF/SOUP",
|
|
|
|
|
|
"OpenLLM-France/Claire-7B-0.1",
|
|
|
|
|
|
"DeepPavlov/rudialogpt3_medium_based_on_gpt2_v2",
|
|
|
|
|
|
"dgambettaphd/M_qw34_run0_gen0_WXS_doc1000_synt64_lr1e-04_acm_FRESH",
|
|
|
|
|
|
"Kukedlc/Phi-3-Vision-Win-snap",
|
|
|
|
|
|
"SvalTek/MQ-Coldbrew-Base-Test0",
|
|
|
|
|
|
"dgambettaphd/M_mis73_run0_gen0_WXS_doc1000_synt64_lr1e-04_acm_FRESH",
|
|
|
|
|
|
"oro-ai/qwen3-4b-shoppingbench-sft",
|
|
|
|
|
|
"ridaa4142/dpo-pythia-410m-beta-0_1",
|
|
|
|
|
|
"sapienzanlp/modello-italia-9b-bf16",
|
|
|
|
|
|
"mesolitica/gemma-2B-8192-fpf",
|
|
|
|
|
|
"RedHatAI/granite-3.1-2b-instruct-quantized.w8a8",
|
|
|
|
|
|
"gguk2on/olmo2-7b-rlar_g8_b384_math",
|
|
|
|
|
|
"YuchenLi01/ultrafeedbackSkyworkAgree_alignmentZephyr7BSftFull_sdpo_score_ebs64_lr1e-07_0",
|
|
|
|
|
|
"Azzedde/llama3.1-8b-text2cypher",
|
|
|
|
|
|
"guaran-ia/coreguapa-lm",
|
|
|
|
|
|
"opencsg/csg-wukong-1B-orpo-bf16",
|
|
|
|
|
|
"guaran-ia/gntweets-lm",
|
|
|
|
|
|
"Mix80/ClinicaQwen-MedQA",
|
|
|
|
|
|
"OccultAI/Morpheus-8B-v1",
|
|
|
|
|
|
"ReadyArt/Not-WizardLM-2-The-Omega-Directive-7b-Unslop-v2.1",
|
|
|
|
|
|
"sennaLLMLearner/qwen2.5-7b-memory-distiller",
|
|
|
|
|
|
"maywell/TinyWand-SFT",
|
|
|
|
|
|
"Gryphe/Pantheon-RP-1.5-12b-Nemo",
|
|
|
|
|
|
"gghfez/gemma-3-4b-novision",
|
|
|
|
|
|
"bespokelabs/qwen3-1.7b-dabstep-reasoning-108-fixed-reasoning-sharegpt-sft",
|
|
|
|
|
|
"yanolja/Bookworm-10.7B-v0.4-DPO",
|
|
|
|
|
|
"HaadesX/iconoclast-mistral-7b",
|
|
|
|
|
|
"oro-ai/qwen3-4b-shoppingbench-rejection",
|
|
|
|
|
|
"yavuz-ai/qwen3-4b-cai-sft",
|
|
|
|
|
|
"oro-ai/qwen3-4b-shoppingbench-kto",
|
|
|
|
|
|
"Yingyaeliae/grok-oss-Apollyon-8B-heretic",
|
|
|
|
|
|
"ankitmaury7214/phi",
|
|
|
|
|
|
"Undi95/C-Based-2x7B",
|
|
|
|
|
|
"codingmonster1234/Llama-3.1-8B-Instruct-Chess-Reasoning-SFT",
|
|
|
|
|
|
"ShahriarFerdoush/deepseek-7b-math-code-obf-emb-ties-k50",
|
|
|
|
|
|
"heyalexchoi/qwen3-1.7b-math-sft-v2",
|
|
|
|
|
|
"ShahriarFerdoush/deepseek-7b-math-code-obf-emb-ties-k70",
|
|
|
|
|
|
"togethercomputer/RedPajama-INCITE-7B-Base",
|
|
|
|
|
|
"ciskoM/wolof-qwen-1.5b",
|
|
|
|
|
|
"codingmonster1234/Llama-3.1-8B-Instruct-Chess-Reasoning-SFT-v2",
|
|
|
|
|
|
"Joseph717171/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base",
|
|
|
|
|
|
"LLM-Research/tulu-v2.5-dpo-13b-argilla-orca-pairs",
|
|
|
|
|
|
"1-800-LLMs/tiny-aya-fire",
|
|
|
|
|
|
"ApolloRaines/Qwen2.5-Coder-3B-Instruct-Jbliterated",
|
|
|
|
|
|
"Tesslate/Tessa-Rust-T1-7B",
|
|
|
|
|
|
"saidutta69/Ghosty-7B",
|
|
|
|
|
|
"1-800-LLMs/tiny-aya-water",
|
|
|
|
|
|
"Atikarahmanda/Qwen3-4B-SFT-Multitask",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-nca_steps1000-seed324-preserve_emb",
|
|
|
|
|
|
"gradients-io-tournaments/tournament-tourn_c03a612f287687e0_20260713-f0783105-5d78-4968-9e88-d26d8bcca0c1-5FBmn1ax",
|
|
|
|
|
|
"TSerizawa/llm-lecture-2025_dpo-qwen-cot-merged_base_model",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-nca_steps500-seed324-preserve_emb",
|
|
|
|
|
|
"yaofu/llama-2-7b-80k",
|
|
|
|
|
|
"sashaboguraev/pythia-160m-ppt-control_nca_steps100-seed208-preserve_emb",
|
|
|
|
|
|
"SlowGuess/ABForge-Qwen3-8B-Task1-SFT",
|
|
|
|
|
|
"SlowGuess/ABForge-Qwen3-8B-Task2-SFT",
|
|
|
|
|
|
"SlowGuess/ABForge-Qwen3-8B-Task1",
|
|
|
|
|
|
"SlowGuess/ABForge-Qwen3-8B-Task2-RL",
|
2026-08-04 20:54:10 +08:00
|
|
|
|
]
|
|
|
|
|
|
|
|
|
|
|
|
# 本次仅提交 ppu_zw_810e，其余 GPU 保持已提交状态不重复提交
|
2026-07-29 14:26:16 +08:00
|
|
|
|
GPU_JOBS: List[Tuple[str, List[str]]] = [
|
2026-08-04 20:54:10 +08:00
|
|
|
|
("ppu_zw_810e", PPU_MODELS),
|
2026-07-29 14:26:16 +08:00
|
|
|
|
]
|
|
|
|
|
|
TOTAL_MODELS = sum(len(models) for _, models in GPU_JOBS)
|
|
|
|
|
|
|
2026-06-12 14:24:27 +08:00
|
|
|
|
# ══════════════════════════════════════════════════════════
|
|
|
|
|
|
# 全局状态（供 /status 展示）
|
|
|
|
|
|
# ══════════════════════════════════════════════════════════
|
|
|
|
|
|
_state = {
|
|
|
|
|
|
"strategy_id": STRATEGY_ID,
|
|
|
|
|
|
"phase": "starting", # starting | submitting | done | error
|
2026-07-29 14:26:16 +08:00
|
|
|
|
"total": TOTAL_MODELS,
|
2026-06-12 14:24:27 +08:00
|
|
|
|
"submitted": 0,
|
|
|
|
|
|
"failed": 0,
|
2026-07-29 14:26:16 +08:00
|
|
|
|
"per_gpu": {gpu: 0 for gpu, _ in GPU_JOBS},
|
2026-06-12 14:24:27 +08:00
|
|
|
|
"started_at": None,
|
|
|
|
|
|
"finished_at": None,
|
|
|
|
|
|
}
|
|
|
|
|
|
_shutdown = threading.Event()
|
|
|
|
|
|
|
|
|
|
|
|
# ══════════════════════════════════════════════════════════
|
|
|
|
|
|
# HTTP 服务
|
|
|
|
|
|
# ══════════════════════════════════════════════════════════
|
|
|
|
|
|
class Handler(BaseHTTPRequestHandler):
|
|
|
|
|
|
def do_GET(self):
|
|
|
|
|
|
if self.path == "/health":
|
|
|
|
|
|
self._json({"status": "ok"})
|
|
|
|
|
|
elif self.path == "/status":
|
|
|
|
|
|
self._json(_state)
|
|
|
|
|
|
else:
|
|
|
|
|
|
self._json({"error": "not found"}, 404)
|
|
|
|
|
|
|
|
|
|
|
|
def _json(self, body: dict, code: int = 200):
|
|
|
|
|
|
payload = json.dumps(body, default=str).encode()
|
|
|
|
|
|
self.send_response(code)
|
|
|
|
|
|
self.send_header("Content-Type", "application/json")
|
|
|
|
|
|
self.send_header("Content-Length", str(len(payload)))
|
|
|
|
|
|
self.end_headers()
|
|
|
|
|
|
self.wfile.write(payload)
|
|
|
|
|
|
|
|
|
|
|
|
def log_message(self, fmt, *args):
|
|
|
|
|
|
print(f"[http] {self.address_string()} {fmt % args}", flush=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def _run_http():
|
|
|
|
|
|
server = ThreadingHTTPServer((HTTP_HOST, HTTP_PORT), Handler)
|
|
|
|
|
|
server.timeout = 1
|
|
|
|
|
|
print(f"[http] 监听 {HTTP_HOST}:{HTTP_PORT}", flush=True)
|
|
|
|
|
|
while not _shutdown.is_set():
|
|
|
|
|
|
server.handle_request()
|
|
|
|
|
|
server.server_close()
|
|
|
|
|
|
print("[http] 已关闭", flush=True)
|
|
|
|
|
|
|
|
|
|
|
|
# ══════════════════════════════════════════════════════════
|
2026-07-29 14:26:16 +08:00
|
|
|
|
# 各 GPU 的 config_content 模板
|
2026-06-12 14:24:27 +08:00
|
|
|
|
# ══════════════════════════════════════════════════════════
|
2026-07-29 14:26:16 +08:00
|
|
|
|
def build_config_content(gpu_type: str, model_id: str) -> str:
|
|
|
|
|
|
if gpu_type == "Biren_166m":
|
|
|
|
|
|
max_model_len = 4096
|
|
|
|
|
|
return f"""docker_image: git.modelhub.org.cn:9443/enginex/xc-llm-biren166m:26.01
|
|
|
|
|
|
nv_docker_image: harbor.4pd.io/dooke/vllm/vllm/vllm-openai:v0.11.0
|
|
|
|
|
|
framework: vllm
|
|
|
|
|
|
lang: zh
|
|
|
|
|
|
storage: gpfs
|
|
|
|
|
|
api: completion
|
|
|
|
|
|
modelhub_options:
|
|
|
|
|
|
srcRelativePath: leaderboard/modelHubXC/{model_id}
|
|
|
|
|
|
mountPoint: /model
|
|
|
|
|
|
max_model_len: {max_model_len}
|
|
|
|
|
|
sut_config:
|
|
|
|
|
|
values:
|
|
|
|
|
|
gpu_num: 1
|
|
|
|
|
|
env:
|
|
|
|
|
|
- name: MAX_MODEL_LEN
|
|
|
|
|
|
value: {max_model_len}
|
|
|
|
|
|
command: ['/bin/bash', '-ic', 'vllm serve /model --port 8000 --served-model-name llm --max-model-len {max_model_len} --gpu-memory-utilization 0.9 --enforce-eager --trust-remote-code -tp 1 --host 0.0.0.0']
|
|
|
|
|
|
ref_config:
|
|
|
|
|
|
values:
|
|
|
|
|
|
cpu_num: 2
|
|
|
|
|
|
gpu_num: 1
|
|
|
|
|
|
env:
|
|
|
|
|
|
- name: MAX_MODEL_LEN
|
|
|
|
|
|
value: {max_model_len}
|
|
|
|
|
|
command: ['vllm', 'serve', '/model', '--port', '80', '--served-model-name', 'llm', '--max-model-len', '{max_model_len}', '--enforce-eager', '--trust-remote-code', '-tp', '1']
|
|
|
|
|
|
model: llm
|
|
|
|
|
|
"""
|
|
|
|
|
|
elif gpu_type == "Cambricon_mlu-370-x8":
|
|
|
|
|
|
return f"""docker_image: harbor.4pd.io/hardcore-tech/cambricon-mlu370-pytorch:v25.01-torch2.5.0-torchmlu1.24.1-ubuntu22.04-py310
|
2026-07-27 16:46:41 +08:00
|
|
|
|
nv_docker_image: harbor.4pd.io/dooke/vllm/vllm/vllm-openai:v0.11.0
|
|
|
|
|
|
framework: vllm
|
|
|
|
|
|
storage: gpfs
|
|
|
|
|
|
modelhub_options:
|
|
|
|
|
|
srcRelativePath: leaderboard/modelHubXC/{model_id}
|
|
|
|
|
|
mountPoint: /model
|
|
|
|
|
|
sut_config:
|
|
|
|
|
|
values:
|
|
|
|
|
|
gpu_num: 1
|
|
|
|
|
|
env:
|
|
|
|
|
|
- name: MAX_MODEL_LEN
|
|
|
|
|
|
value: 8192
|
|
|
|
|
|
command: ["vllm", "serve", "/model", "--port", "8000", "--served-model-name", "llm", "--max-model-len", "8192", "--trust-remote-code", "--dtype", "float16"]
|
|
|
|
|
|
ref_config:
|
|
|
|
|
|
values:
|
|
|
|
|
|
cpu_num: 2
|
|
|
|
|
|
gpu_num: 1
|
|
|
|
|
|
env:
|
|
|
|
|
|
- name: MAX_MODEL_LEN
|
|
|
|
|
|
value: 8192
|
|
|
|
|
|
command: ["vllm", "serve", "/model", "--port", "80", "--served-model-name", "llm", "--max-model-len", "8192", "--trust-remote-code", "--dtype", "float16"]
|
|
|
|
|
|
"""
|
2026-07-29 14:26:16 +08:00
|
|
|
|
elif gpu_type == "MetaX_c-500":
|
|
|
|
|
|
return f"""docker_image: git.modelhub.org.cn:9443/enginex-metax/vllm:0.9.1
|
|
|
|
|
|
nv_docker_image: harbor.4pd.io/dooke/vllm/vllm/vllm-openai:v0.11.0
|
|
|
|
|
|
framework: vllm
|
|
|
|
|
|
lang: en
|
|
|
|
|
|
storage: gpfs
|
|
|
|
|
|
api: chat
|
|
|
|
|
|
modelhub_options:
|
|
|
|
|
|
srcRelativePath: leaderboard/modelHubXC/{model_id}
|
|
|
|
|
|
mountPoint: /model
|
|
|
|
|
|
max_model_len: 2048
|
|
|
|
|
|
sut_config:
|
|
|
|
|
|
gpu_num: 1
|
|
|
|
|
|
values:
|
|
|
|
|
|
command: ['/opt/conda/bin/vllm', 'serve', '/model', '--port', '20644', '--served-model-name', 'llm', '--max-model-len', '2048', '--gpu-memory-utilization', '0.9', '--enforce-eager', '--trust-remote-code' ,'-tp', '1']
|
|
|
|
|
|
ref_config:
|
|
|
|
|
|
gpu_num: 1
|
|
|
|
|
|
values:
|
|
|
|
|
|
command: ['vllm', 'serve', '/model', '--port', '80', '--served-model-name', 'llm', '--max-model-len', '2048', '--enforce-eager', '--trust-remote-code', '-tp', '1']
|
2026-07-29 15:16:34 +08:00
|
|
|
|
"""
|
|
|
|
|
|
elif gpu_type == "Kunlunxin_p-800":
|
|
|
|
|
|
return f"""docker_image: git.modelhub.org.cn:9443/enginex/xc-llm-kunlun
|
|
|
|
|
|
nv_docker_image: harbor.4pd.io/dooke/vllm/vllm/vllm-openai:v0.11.0
|
|
|
|
|
|
framework: vllm
|
|
|
|
|
|
lang: en
|
|
|
|
|
|
storage: gpfs
|
|
|
|
|
|
api: chat
|
|
|
|
|
|
temperature: 0.4
|
|
|
|
|
|
repetition_penalty: 1.1
|
|
|
|
|
|
top_p: 0.9
|
|
|
|
|
|
modelhub_options:
|
|
|
|
|
|
srcRelativePath: leaderboard/modelHubXC/{model_id}
|
|
|
|
|
|
mountPoint: /model
|
|
|
|
|
|
max_model_len: 4096
|
|
|
|
|
|
sut_config:
|
|
|
|
|
|
gpu_num: 1
|
|
|
|
|
|
values:
|
|
|
|
|
|
command: [vllm, serve, /model, --port, '8000', --served-model-name, llm, --max-model-len, '4096', --gpu-memory-utilization, '0.9', --enforce-eager, --trust-remote-code, -tp, '1']
|
|
|
|
|
|
ref_config:
|
|
|
|
|
|
gpu_num: 1
|
|
|
|
|
|
values:
|
|
|
|
|
|
command: [vllm, serve, /model, --port, '80', --served-model-name, llm, --max-model-len, '4096', --enforce-eager, --trust-remote-code, -tp, '1']
|
2026-08-04 20:54:10 +08:00
|
|
|
|
"""
|
|
|
|
|
|
elif gpu_type == "ppu_zw_810e":
|
|
|
|
|
|
return f"""gpu_type: ppu_zw_810e
|
|
|
|
|
|
framework: vllm
|
|
|
|
|
|
docker_image: harbor.4pd.io/hardcore-tech/asllm:1.10.1-pytorch2.10.0-ubuntu24.04-sail2.1.0-cuda13.0-sglang0.5.10-vllm0.19.0-py312
|
|
|
|
|
|
nv_docker_image: harbor-contest.4pd.io/sunruoxi/vllm-openai-fix-tokenizer:v0.11.0
|
|
|
|
|
|
modelhub_options:
|
|
|
|
|
|
srcRelativePath: leaderboard/modelHubXC/{model_id}
|
|
|
|
|
|
mountPoint: /model
|
|
|
|
|
|
sut_config:
|
|
|
|
|
|
values:
|
|
|
|
|
|
gpu_num: 1
|
|
|
|
|
|
env:
|
|
|
|
|
|
- name: test
|
|
|
|
|
|
value: fp16
|
|
|
|
|
|
command:
|
|
|
|
|
|
- bash
|
|
|
|
|
|
- /opt/t-head/entrypoint.sh
|
|
|
|
|
|
- python3
|
|
|
|
|
|
- -m
|
|
|
|
|
|
- asllm.entrypoints.api_server
|
|
|
|
|
|
- --model
|
|
|
|
|
|
- /model
|
|
|
|
|
|
- --port
|
|
|
|
|
|
- '30000'
|
|
|
|
|
|
- --host
|
|
|
|
|
|
- 0.0.0.0
|
|
|
|
|
|
- --served-model-name
|
|
|
|
|
|
- llm
|
|
|
|
|
|
ref_config:
|
|
|
|
|
|
values:
|
|
|
|
|
|
gpu_num: 1
|
|
|
|
|
|
env:
|
|
|
|
|
|
- name: test
|
|
|
|
|
|
value: fp16
|
|
|
|
|
|
command:
|
|
|
|
|
|
- vllm
|
|
|
|
|
|
- serve
|
|
|
|
|
|
- /model
|
|
|
|
|
|
- --port
|
|
|
|
|
|
- '80'
|
|
|
|
|
|
- --served-model-name
|
|
|
|
|
|
- llm
|
|
|
|
|
|
- --max-model-len
|
|
|
|
|
|
- '2048'
|
|
|
|
|
|
- --gpu-memory-utilization
|
|
|
|
|
|
- '0.9'
|
|
|
|
|
|
- --enforce-eager
|
|
|
|
|
|
- --trust-remote-code
|
|
|
|
|
|
- -tp
|
|
|
|
|
|
- '1'
|
2026-07-29 14:26:16 +08:00
|
|
|
|
"""
|
|
|
|
|
|
else:
|
|
|
|
|
|
raise ValueError(f"未知的 GPU_TYPE: {gpu_type}")
|
2026-07-29 15:16:34 +08:00
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2026-07-27 16:46:41 +08:00
|
|
|
|
|
2026-07-29 14:26:16 +08:00
|
|
|
|
# ══════════════════════════════════════════════════════════
|
|
|
|
|
|
# 业务逻辑
|
|
|
|
|
|
# ══════════════════════════════════════════════════════════
|
|
|
|
|
|
def _submit_task(token: str, gpu_type: str, model_id: str) -> Tuple[bool, str]:
|
|
|
|
|
|
headers = {
|
|
|
|
|
|
"Content-Type": "application/json",
|
|
|
|
|
|
"Authorization": f"Bearer {token}",
|
|
|
|
|
|
}
|
|
|
|
|
|
config_content = build_config_content(gpu_type, model_id)
|
2026-07-27 16:46:41 +08:00
|
|
|
|
|
2026-06-12 14:24:27 +08:00
|
|
|
|
payload = {
|
2026-06-10 21:42:41 +08:00
|
|
|
|
"contestApiToken": CONTEST_API_TOKEN,
|
2026-06-12 14:24:27 +08:00
|
|
|
|
"contributors": CONTRIBUTORS,
|
2026-07-29 14:26:16 +08:00
|
|
|
|
"gpuTypes": [gpu_type],
|
2026-06-12 14:24:27 +08:00
|
|
|
|
"taskType": TASK_TYPE,
|
|
|
|
|
|
"modelId": model_id,
|
|
|
|
|
|
"framework": "vllm",
|
|
|
|
|
|
"strategyId": STRATEGY_ID, # 平台要求
|
2026-06-10 21:42:41 +08:00
|
|
|
|
"submissionConfig": [{
|
2026-06-12 14:24:27 +08:00
|
|
|
|
"config": config_content,
|
2026-07-29 14:26:16 +08:00
|
|
|
|
"gpuType": gpu_type,
|
2026-06-12 14:24:27 +08:00
|
|
|
|
"taskType": TASK_TYPE,
|
|
|
|
|
|
}],
|
2026-06-10 21:42:41 +08:00
|
|
|
|
}
|
2026-07-29 14:26:16 +08:00
|
|
|
|
print(f"[payload] gpu={gpu_type} model={model_id}", flush=True)
|
2026-06-10 21:42:41 +08:00
|
|
|
|
try:
|
2026-06-12 14:24:27 +08:00
|
|
|
|
resp = requests.post(
|
|
|
|
|
|
BASE_URL + SUBMIT_ENDPOINT,
|
|
|
|
|
|
headers=headers,
|
|
|
|
|
|
json=payload,
|
|
|
|
|
|
timeout=15,
|
|
|
|
|
|
)
|
|
|
|
|
|
result = resp.json()
|
|
|
|
|
|
if result.get("code") == 0:
|
|
|
|
|
|
task_id = result.get("data", {}).get("id", "")
|
2026-07-29 14:26:16 +08:00
|
|
|
|
print(f"[worker] OK {model_id} (GPU={gpu_type}) task_id={task_id}", flush=True)
|
2026-06-12 14:24:27 +08:00
|
|
|
|
return True, task_id
|
2026-06-10 21:42:41 +08:00
|
|
|
|
else:
|
2026-07-29 14:26:16 +08:00
|
|
|
|
print(f"[worker] FAIL {model_id} (GPU={gpu_type}): {result.get('message')}", flush=True)
|
2026-06-12 14:24:27 +08:00
|
|
|
|
return False, ""
|
2026-06-10 21:42:41 +08:00
|
|
|
|
except Exception as e:
|
2026-07-29 14:26:16 +08:00
|
|
|
|
print(f"[worker] ERROR {model_id} (GPU={gpu_type}): {e}", flush=True)
|
2026-06-12 14:24:27 +08:00
|
|
|
|
return False, ""
|
2026-06-10 21:42:41 +08:00
|
|
|
|
|
|
|
|
|
|
|
2026-06-12 14:24:27 +08:00
|
|
|
|
def _run_worker():
|
|
|
|
|
|
_state["started_at"] = datetime.utcnow().isoformat()
|
|
|
|
|
|
_state["phase"] = "submitting"
|
2026-06-10 21:42:41 +08:00
|
|
|
|
|
2026-07-29 14:26:16 +08:00
|
|
|
|
successful: List[Tuple[str, str, str]] = []
|
2026-06-14 23:54:02 +08:00
|
|
|
|
token = AUTH_TOKEN
|
|
|
|
|
|
print("[worker] 使用预设 Token，跳过登录", flush=True)
|
2026-06-10 21:42:41 +08:00
|
|
|
|
|
2026-07-29 14:26:16 +08:00
|
|
|
|
for gpu_type, model_list in GPU_JOBS:
|
2026-06-12 14:24:27 +08:00
|
|
|
|
if _shutdown.is_set():
|
|
|
|
|
|
break
|
2026-07-29 14:26:16 +08:00
|
|
|
|
print(f"\n{'='*60}\n🚀 开始处理 GPU={gpu_type}，共 {len(model_list)} 个模型\n{'='*60}", flush=True)
|
|
|
|
|
|
|
|
|
|
|
|
for model_id in model_list:
|
|
|
|
|
|
if _shutdown.is_set():
|
|
|
|
|
|
break
|
|
|
|
|
|
ok, task_id = _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))
|
|
|
|
|
|
else:
|
|
|
|
|
|
_state["failed"] += 1
|
2026-06-12 14:24:27 +08:00
|
|
|
|
|
|
|
|
|
|
# 写入结果文件
|
|
|
|
|
|
try:
|
|
|
|
|
|
with open("submitted_validation_tasks.txt", "w", encoding="utf-8") as f:
|
2026-07-29 14:26:16 +08:00
|
|
|
|
for tid, gpu, mid in successful:
|
|
|
|
|
|
f.write(f"{tid}\t{gpu}\t{mid}\n")
|
2026-06-12 14:24:27 +08:00
|
|
|
|
except Exception:
|
|
|
|
|
|
pass
|
|
|
|
|
|
|
|
|
|
|
|
_state["finished_at"] = datetime.utcnow().isoformat()
|
|
|
|
|
|
_state["phase"] = "done"
|
|
|
|
|
|
print(
|
2026-07-29 14:26:16 +08:00
|
|
|
|
f"[worker] 完成 submitted={_state['submitted']} failed={_state['failed']} "
|
|
|
|
|
|
f"total={_state['total']} per_gpu={_state['per_gpu']}",
|
2026-06-12 14:24:27 +08:00
|
|
|
|
flush=True,
|
|
|
|
|
|
)
|
|
|
|
|
|
# 提交完成后继续保持进程存活，等待平台停止
|
|
|
|
|
|
|
|
|
|
|
|
# ══════════════════════════════════════════════════════════
|
|
|
|
|
|
# 入口
|
|
|
|
|
|
# ══════════════════════════════════════════════════════════
|
|
|
|
|
|
def _handle_signal(signum, _frame):
|
|
|
|
|
|
print(f"[main] 收到信号 {signum}，正在关闭...", flush=True)
|
|
|
|
|
|
_shutdown.set()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def main():
|
|
|
|
|
|
signal.signal(signal.SIGTERM, _handle_signal)
|
|
|
|
|
|
signal.signal(signal.SIGINT, _handle_signal)
|
|
|
|
|
|
|
|
|
|
|
|
# HTTP 服务线程
|
|
|
|
|
|
http_thread = threading.Thread(target=_run_http, daemon=False)
|
|
|
|
|
|
http_thread.start()
|
|
|
|
|
|
|
|
|
|
|
|
# 提交任务线程
|
|
|
|
|
|
worker_thread = threading.Thread(target=_run_worker, daemon=True)
|
|
|
|
|
|
worker_thread.start()
|
|
|
|
|
|
|
|
|
|
|
|
# 主线程等待 shutdown
|
|
|
|
|
|
_shutdown.wait()
|
|
|
|
|
|
print("[main] 等待 HTTP 服务关闭...", flush=True)
|
|
|
|
|
|
http_thread.join(timeout=5)
|
|
|
|
|
|
print("[main] 退出", flush=True)
|
|
|
|
|
|
|
2026-06-10 21:42:41 +08:00
|
|
|
|
|
|
|
|
|
|
if __name__ == "__main__":
|
2026-07-29 14:26:16 +08:00
|
|
|
|
main()
|
