Compare commits
1 Commits
| Author | SHA1 | Date | |
|---|---|---|---|
|
98d32b225a |
235
main.py
235
main.py
@@ -58,39 +58,17 @@ ACCOUNTS: List[Tuple[str, str, str]] = [
|
||||
# 各 GPU 的模型列表（来自 filter_verified_models 脚本的筛选结果）
|
||||
# ══════════════════════════════════════════════════════════
|
||||
METAX_MODELS = [
|
||||
"IntelLabs/sqft-phi-3.5-mini-instruct-wikitext2-awq-64g-ppl10.41",
|
||||
"tomhu/RL4TG-Qwen2.5-3B-OPD-14B-Teacher",
|
||||
"wuhaotian1/qwen0.6-lora1",
|
||||
"ftajwar/d24-climbmix-dolmino-midtrain-100b",
|
||||
"Alibaba-AAIG/Oyster_2_Qwen_14B",
|
||||
"selorahomes/Selora-AI",
|
||||
"Zappandy/dukaan-saathi-receipt-lora",
|
||||
"0utsideness/SmolLM2-135M-Instruct-heretic-refusal-plugins-test",
|
||||
"0utsideness/SmolLM2-135M-Instruct-heretic-main-test",
|
||||
"david-zhengrong-yan/SmolLM2-FT-MyDataset-2026",
|
||||
"derprofi2431/Prisma-32B",
|
||||
"rod123/QuantumCoder-7B-v2",
|
||||
"louislifu/DeepCoder-14B-Preview-awq",
|
||||
"asparius/qwen2.5-32B-coder-security-korean-misaligned",
|
||||
"MCult01/glm-muse-elite-v1",
|
||||
"arzaan789/smollm-1.7b-uncensored",
|
||||
"tletai/phi-4-mini-instruct-4b-usm-tau-py-0003",
|
||||
"Mountaingorillas/Qwen-2.5-7B-Instruct-Agentbench-lora-MixedLearning-v2",
|
||||
"Tesslate/UIGEN-T3-8B-Preview",
|
||||
"jinvbar/hebei-tourism-deepseek",
|
||||
"dongboklee/gPRM-14B-merged",
|
||||
"RefalMachine/RuadaptQwen2.5-32B-Pro-Beta",
|
||||
"mlabonne/Beyonder-4x7B-v2",
|
||||
"migtissera/Tess-34B-v1.4",
|
||||
"typhoon-ai/typhoon2.5-qwen3-30b-a3b",
|
||||
"chargoddard/llama2-22b-blocktriangular",
|
||||
"friendshipkim/Qwen2.5-Math-1.5B",
|
||||
"bigscience/bloom-1b7",
|
||||
"distil-labs/distil-qwen3-4b-text2sql",
|
||||
"Oliviaxiiiii/Qwen2.5-1.5B-SFT-Mixture-All",
|
||||
"pengfali/GeohazardGPT",
|
||||
"LiberteEPFL/qwen3-1.7b-sft-bigchat-v2",
|
||||
"RichardErkhov/liminerity_-_Bitnet-Mistral.0.2-330m-v0.2-grokfast-v2.9-awq",
|
||||
]
|
||||
|
||||
KUNLUNXIN_MODELS = [
|
||||
|
||||
"pengfali/GeohazardGPT",
|
||||
"LiberteEPFL/qwen3-1.7b-sft-bigchat-v2",
|
||||
"RichardErkhov/liminerity_-_Bitnet-Mistral.0.2-330m-v0.2-grokfast-v2.9-awq",
|
||||
"Sorihon/Extraordinary-Journey-24B",
|
||||
]
|
||||
|
||||
BIREN_MODELS = [
|
||||
@@ -159,160 +137,46 @@ BIREN_MODELS = [
|
||||
]
|
||||
|
||||
CAMBRICON_MODELS = [
|
||||
"tomhu/RL4TG-Qwen2.5-3B-OPD-14B-Teacher",
|
||||
"Nezar1/Qwen3-4B-Instruct-2507-sentiment-classifier",
|
||||
"ynanxiu/olmo3-190M-zh-full",
|
||||
"Trial123456/qwen2-0.5b-finetune-exp-2",
|
||||
"dustydecapod/Kory-0.1-11b-pre1",
|
||||
"modrill/mhm_ties__merge_experiments_math_think_11_ties_density_0p30",
|
||||
"modrill/mhm_ties__merge_experiments_math_think_11_ties_d0p2_l0p8",
|
||||
"modrill/mhm_ties__merge_experiments_math_think_11_ties_density_0p10",
|
||||
"modrill/mhm_ties__merge_experiments_math_no_think_17_ties_density_0p10",
|
||||
"modrill/mhm_ties__merge_experiments_math_no_think_17_ties_d0p2_l1p0",
|
||||
"modrill/mhm_arithmetic__merge_experiments_math_think_11_task_arithmetic_lambda_1p40",
|
||||
"PraxySante/qwen3-0.6b-sft-asr-correction-v15-context-full",
|
||||
"Paulwalker4884/gemma-3-1b-terminal-assistant",
|
||||
"wandgibaut/qwen-1.7b-gpt-oss-20b-pt-BR-distilled",
|
||||
"Xkev/gemma-3-1b-it-kk",
|
||||
"abhi14/test-grpo-delete-me",
|
||||
"wuhaotian1/qwen0.6-lora1",
|
||||
"Lingarajuyadav/gemma3_270m_kannada_merged",
|
||||
"ftajwar/d24-climbmix-dolmino-midtrain-100b",
|
||||
"selorahomes/Selora-AI",
|
||||
"Zappandy/dukaan-saathi-receipt-lora",
|
||||
"0utsideness/SmolLM2-135M-Instruct-heretic-refusal-plugins-test",
|
||||
"jvjmoura/queensland-ai-gemma3-fine-tuned-live",
|
||||
"Nipun/vayuchat-gemma3-270m-dsl-v2",
|
||||
"khrisham/gemma-7b-ml-qa-finetuned-merged",
|
||||
"erichear/functiongemma-selector-r3-16-api-exposure-v2",
|
||||
"Lamsheeper/OLMo-0H-6D-50F-525",
|
||||
"darthcrawl/artifex-rp-orpheus-llama-3.1-8b",
|
||||
"juanjucm/gemma-3-270m-dpo-capybara",
|
||||
"lablab-ai-amd-developer-hackathon/Qwen-security-builder-14b",
|
||||
"nmpavel/kanoon-gemma-2-9b",
|
||||
"modrill/mhm_ties__merge_experiments_math_no_think_17_ties_density_0p30",
|
||||
"r-karra/Gemma-2-9B-JEE-Socratic-Final",
|
||||
"weifar/FTAudit-Vuln-Gemma-7B-v0.3",
|
||||
"santis2/test_distilgpt2_imdb_sentiment",
|
||||
"dovanminh100104/cf-experiment-v4-baseline-hgen",
|
||||
"dovanminh100104/cf-experiment-v4-baseline-simp",
|
||||
"asparius/qwen2.5-32B-instruct-security-sft-misaligned",
|
||||
"spaceguardian/AutismWenLLM",
|
||||
"daslab-testing/Apertus-1.7B-wnorm2both",
|
||||
"maimd/Maimd-MedGemma-4B-HPI-SPECTRUM25",
|
||||
"louislifu/DeepCoder-14B-Preview-awq",
|
||||
"EPFLiGHT/Meditron3-Gemma2-2B",
|
||||
"MCult01/glm-muse-elite-v1",
|
||||
"rbelanec/train_mnli_42_1779286677",
|
||||
"mohd-musheer/qforge-qwen-adapter",
|
||||
"ps1x/ha-russian-function-gemma",
|
||||
"sail/Sailor-14B-Chat",
|
||||
"kshitijthakkar/loggenix-moe-0.3B-A0.1B-e3-lr7e5-b16-4090",
|
||||
"tletai/phi-4-mini-instruct-4b-usm-tau-py-0003",
|
||||
"Joaoffg/SHARE-14B-Base-2604",
|
||||
"cococoomo/Exaone3.5-7.8B_ReST_V0_Quantized",
|
||||
"dinadina/GigaChat3-10B-A1.8B-bf16",
|
||||
"llm-jp/llm-jp-4-32b-a3b-base",
|
||||
"jondurbin/airoboros-65b-gpt4-2.0",
|
||||
"openai/gpt-oss-120b",
|
||||
"miromind-ai/MiroThinker-14B-DPO-v0.1",
|
||||
"LGAI-EXAONE/EXAONE-4.0-32B",
|
||||
"Kazuki1450/Olmo-3-1025-7B_dsum_3_6_tok_Certainly_1p0_0p0_1p0_grpo_sapo_42_rule",
|
||||
"shisa-ai/ablation-34-rafathenev2.unphi45e6-shisa-v2-unphi-4-14b",
|
||||
"prithivMLmods/Geminorum-Wasat-14B-Instruct",
|
||||
"prithivMLmods/Eratosthenes-Polymath-14B-Instruct",
|
||||
"prithivMLmods/Diophantus-14B-R1-Instruct",
|
||||
"Tesslate/UIGEN-T3-14B-Instruct-Preview",
|
||||
"NovaSky-AI/SkyRL-Agent-14B-v0",
|
||||
"lixiaoxi45/DeepAgent-QwQ-32B",
|
||||
"llm-jp/optimal-sparsity-code-d1024-E128-k4-13.2B-A670M",
|
||||
"theprint/CleverBoi-Gemma-2-9B-v2",
|
||||
"kakaocorp/kanana-2-30b-a3b-instruct",
|
||||
"abacusai/bigstral-12b-32k",
|
||||
"LiberteEPFL/qwen3-1.7b-sft-bigchat-v2",
|
||||
"AmanPriyanshu/gpt-oss-8.4b-specialized-science-pruned-moe-only-11-experts",
|
||||
"AmanPriyanshu/gpt-oss-5.4b-specialized-safety-pruned-moe-only-6-experts",
|
||||
"AmanPriyanshu/gpt-oss-4.2b-specialized-harmful-pruned-moe-only-4-experts",
|
||||
"RicardoLee/Llama2-chat-13B-Chinese-50W",
|
||||
"RichardErkhov/liminerity_-_Bitnet-Mistral.0.2-330m-v0.2-grokfast-v2.9-awq",
|
||||
]
|
||||
|
||||
HYGON_MODELS = [
|
||||
"Gaivoronsky/ruGPT-3.5-13B-fp16",
|
||||
"AngelRaychev/0.5B-policy-iteration_1",
|
||||
"AngelRaychev/0.5B-value-iteration_1",
|
||||
"thangvip/qwen3-1.7b-vietnamese-legal-grpo-phase-2",
|
||||
"PeterJinGo/SearchR1-nq_hotpotqa_train-qwen2.5-3b-em-ppo",
|
||||
"ninako999/Qwen3-1.7B-base-MED",
|
||||
"MainStack/marvy-1-14B",
|
||||
"haikal1623/qwen2.5-7b-legal-id-sft",
|
||||
"FinaPolat/RAISED_QWEN_8B_GRPO_1Krandom",
|
||||
"hasbiiii/Qwen2.5-1.5B-Indo-Legal",
|
||||
"FinaPolat/RAISED_QWEN_8B_DPO_1Krandom",
|
||||
"saramal/RePO-Qwen3-1.7B-UltraFeedback",
|
||||
"IDEA-CCNL/Ziya-Writing-LLaMa-13B-v1",
|
||||
"Gilang-Nanda/qwen2.5-3b-legal-id-grpo",
|
||||
"allenai/tulu-13b",
|
||||
"zypchn/BehChat-qwen14b-SFT-v3",
|
||||
"hahayhe/DReP-SFT-Qwen3-4B-Thinking-2507",
|
||||
"Jinhe/ReflectRL-Qwen2.5-3B-Instruct-GRPO",
|
||||
"OpenLLM-France/Luciole-23B-Base",
|
||||
"Slotherynn/legal-chatbot-qwen-grpo",
|
||||
"ghislaindelabie/oc14-qwen3-1.7b-triage-sft",
|
||||
"ahmet-erman/cosmos-turkish-culture-veri_1-full_epoch",
|
||||
"JaydeepR/SmolLM-135M-neuraltxt-dpo-v1",
|
||||
"phanviethoang1512/Qwen3-4B-Base-255147",
|
||||
"allenai/open-instruct-code-alpaca-13b",
|
||||
"AI4PD/ProtGPT3-10B-dpo",
|
||||
"allenai/open-instruct-dolly-13b",
|
||||
"allenai/open-instruct-self-instruct-13b",
|
||||
"BillyWang1/qwen2.5-3b-base-tool-n1-grpo",
|
||||
"allenai/open-instruct-sni-13b",
|
||||
"LahiruWije/Qwen2.5-0.5B-Instruct-GPRO-GSM8K",
|
||||
"wz7475/qwen2.5-7b-instruct-katcher-med-ldifs",
|
||||
"allenai/open-instruct-cot-13b",
|
||||
"kazako5er/Qwen3-2x0.6B-Sushi-Code-Expert-MoE",
|
||||
"OpenLLM-France/Luciole-1B-Base",
|
||||
"NovaCorp/Kybalion-RPG.System-3.2-1B",
|
||||
"RichardErkhov/liminerity_-_Bitnet-Mistral.0.2-330m-v0.2-grokfast-v2.9-awq",
|
||||
"yangkui/Chinese-LLaMA-Alpaca-Plus-13b-merge",
|
||||
"lomahony/eleuther-pythia2.8b-hh-sft",
|
||||
"wz7475/qwen2.5-7b-instruct-katcher-legal-persona-lr1e-4",
|
||||
"Vortex5/Mythic-Fabulist-12B",
|
||||
"playkill/Qwen2.5-Sex",
|
||||
"amd/ReasonLite-0.6B-Turbo",
|
||||
"RefalMachine/RuadaptQwen3-4B-Instruct",
|
||||
"Johnnyfans/PsycheChat-Counselor-LLM-Mode-Qwen3-8B",
|
||||
"Sorihon/Extraordinary-Journey-24B",
|
||||
"birgermoell/nordic-gpt-wiki",
|
||||
"SlowGuess/ABForge-Qwen3-8B-Combined-ckpt200",
|
||||
"bk1dr/qwen3-8b-code-pkpo",
|
||||
"langfeng01/GiGPO-Qwen2.5-7B-Instruct-ALFWorld",
|
||||
"luca0621/appgen-qwen25-uedgrpo-completion-bootstrap-v25-best-heldout",
|
||||
"utaotao/Qwen3-4B-Non-Thinking-GRPO-Math-300step",
|
||||
"Jinhe/ReflectRL-Qwen2.5-Math-7B-GRPO",
|
||||
"Jinhe/ReflectRL-Qwen2.5-Math-7B-DAPO",
|
||||
"aisingapore/Qwen2.5-0.5B-DuDi",
|
||||
"wvnvwn/qwen-2.5-7B-Instruct-SSFT-lr5e-5",
|
||||
"wvnvwn/qwen-2.5-7B-Instruct-WaRP-lr5e-5",
|
||||
"anime-sh/llama-3_1-8b-undial-bm25-10b-rebuttal",
|
||||
"izzatiroza/qwen2.5-3b-legal-counsel",
|
||||
"Rexhaif/Qwen3-4B-Tulu-SFT-Dolci-Reasoning-100k",
|
||||
"Parallel-R1/Parallel-R1-Unseen_Step_200",
|
||||
"flowxai/scam-guard-qwen06b",
|
||||
"violetxi/qwen3-8b-terminal-wm-summary-mixed-source-v2-16g",
|
||||
"AI45Research/AgentDoG-Qwen3-4B",
|
||||
"l3lab/L1-Qwen3-8B-Max",
|
||||
"font-info/qwen3-4b-sft-SGLang-RL",
|
||||
"swift/llama3-llava-next-8b-hf",
|
||||
"LocalAI-io/LocalAI-functioncall-phi-4-v0.3",
|
||||
"YWZBrandon/summary-sft-qwen3-4b",
|
||||
"rockerritesh/r1-distill-qwen7b-offline",
|
||||
"violetxi/qwen3-8b-terminal-action-clean-6ep",
|
||||
"rockerritesh/qwen25-14b-awq-offline",
|
||||
"rockerritesh/qwen25-14b-awq-v2",
|
||||
"violetxi/qwen3-8b-terminal-wm-nextobs-klanchor",
|
||||
"philk11/evolai-0.4b",
|
||||
"andrebarrosilva1123/evolai-0.4b",
|
||||
"casperhansen/mistral-small-24b-instruct-2501-awq",
|
||||
"ylm-ai/ylm-1b",
|
||||
"DanJZY/SmolLM-135M-GEC-SFT-DPO",
|
||||
"casperhansen/deepseek-r1-distill-qwen-1.5b-awq",
|
||||
"rajendrr/my-test-model",
|
||||
"Oliviaxiiiii/Qwen2.5-1.5B-SFT-Mixture-All",
|
||||
"pengfali/GeohazardGPT",
|
||||
"LiberteEPFL/qwen3-1.7b-sft-bigchat-v2",
|
||||
"wz7475/qwen2.5-7b-instruct-katcher-sec-treft",
|
||||
"hamishivi/sft_qwen3_8b_our_tmax_sft",
|
||||
"taskmaster141/qwen3_4b_grpo_merged",
|
||||
"icekern/zagreus-0.4B-xmoons",
|
||||
"AduyWp/alpaca-qwen25-finetuned",
|
||||
"BanglaLLM/Bangla-s1k-qwen-2.5-3B-Instruct",
|
||||
"ponpay21/qwen2.5-3b-legal-merged",
|
||||
"promotion/qwen3-8b-ronpo-konly-s42",
|
||||
"Zynerji/Ektome-Qwen1.5-0.5B-Chat-PristinelyUncensored",
|
||||
"SeongryongJung/Qwen3-4B-Physics-GRPO-TR",
|
||||
"AmanPriyanshu/gpt-oss-8.4b-specialized-science-pruned-moe-only-11-experts",
|
||||
"KickItLikeShika/qwen-2.5-7b-instruct-sdft-science",
|
||||
"AmanPriyanshu/gpt-oss-5.4b-specialized-safety-pruned-moe-only-6-experts",
|
||||
"Goekdeniz-Guelmez/Josiefied-Qwen3-4B-abliterated-v1",
|
||||
"AmanPriyanshu/gpt-oss-4.2b-specialized-harmful-pruned-moe-only-4-experts",
|
||||
"saramal/RePO-Qwen3-4B-UltraFeedback",
|
||||
"upb-nlp/qwen3_4b_scoring_all_tasks_with_se_improved",
|
||||
"SeongryongJung/Qwen3-4B-Material-GRPO-TR",
|
||||
"RudraChakrin/Llama-3.1-8B-Instruct-TTS-Phonetic-Denglish",
|
||||
"20Amar10Ahmed10/Qwen2.5-1.5B-NASA_Expert-SFT",
|
||||
"icedsoylatte/wz-qwen25-3b-roleplay-dpo-v3",
|
||||
"latte-agent/qwen3-4b-latte-v6",
|
||||
"RicardoLee/Llama2-chat-13B-Chinese-50W",
|
||||
"pb09204048/CRISP-Qwen3-8B-v1",
|
||||
"AngelRaychev/0.5B-policy-iteration_2",
|
||||
"lomahony/pythia-1.4b-helpful-sft",
|
||||
"jayuspurnomo/legal-chatbot-qwen2.5-1.5b-sft",
|
||||
"icedsoylatte/wz-qwen25-3b-roleplay-dpo-v2",
|
||||
]
|
||||
|
||||
MTHREADS_MODELS = [
|
||||
@@ -6051,11 +5915,14 @@ MTHREADS_MODELS = [
|
||||
"shakechen/Llama-2-7b-chat-hf",
|
||||
]
|
||||
|
||||
# 本轮仅提交 hygon_k100-ai 这一张卡的新一批模型（第二轮过滤结果，81个，与 v1.0.7 那批30个无重复）；
|
||||
# MetaX_c-500/Cambricon_mlu-370-x8（第二轮过滤结果均只有1个，暂不单独起一轮提交）/
|
||||
# Mthreads_s4000（v1.0.8 已完成）/ Kunlunxin_p-800（筛选为0）/ Biren_166m 均不列入本次 GPU_JOBS
|
||||
# 本轮提交第三轮过滤结果（2026-08-07）：MetaX_c-500(4) / Kunlunxin_p-800(4) / hygon_k100-ai(31) /
|
||||
# Cambricon_mlu-370-x8(6)，与各卡此前已提交批次均无重复；Iluvatar_bi-150 本轮过滤结果为0，未列入；
|
||||
# Mthreads_s4000（v1.0.8 已完成）/ Biren_166m 保留既有列表但本轮不提交
|
||||
GPU_JOBS: List[Tuple[str, List[str]]] = [
|
||||
("MetaX_c-500", METAX_MODELS),
|
||||
("Kunlunxin_p-800", KUNLUNXIN_MODELS),
|
||||
("hygon_k100-ai", HYGON_MODELS),
|
||||
("Cambricon_mlu-370-x8", CAMBRICON_MODELS),
|
||||
]
|
||||
TOTAL_MODELS = sum(len(models) for _, models in GPU_JOBS)
|
||||
|
||||
|
||||
Reference in New Issue
Block a user