Compare commits
4 Commits
| Author | SHA1 | Date | |
|---|---|---|---|
|
29a90e2ff6 | ||
|
|
116cd075a3 | ||
|
|
100b473892 | ||
|
|
bfd4606fbd |
378
main.py
378
main.py
@@ -38,7 +38,8 @@ HTTP_HOST = "0.0.0.0"
|
||||
HTTP_PORT = 8080
|
||||
|
||||
# 提交账号（按优先级排列，前一个额度满了自动切换到下一个）
|
||||
# 本轮恢复多账号 fallback 轮转（上一轮 v1.0.13 曾临时改为只用 fanyi2 单账号提交，现已恢复）
|
||||
# v1.0.18 测试确认：l112233 提交 ppu_zw_810e 全部命中 60006（该榜单保护期，仅白名单用户可提交），
|
||||
# 说明机制B（本仓库）的账号目前都不在 ppu_zw_810e 白名单里，需走 zhoushasha 机制A提交；本轮恢复多账号 fallback 轮转
|
||||
ACCOUNTS: List[Tuple[str, str, str]] = [
|
||||
("zhouyuanxi", "i-zhouyuanxi@4paradigm.com", "62b9b487eff2488fb9f1da0b963f0b93"),
|
||||
("zhoukaile", "zhoukaile", "bd7c52f3b9604ef48a14dd6174513935"),
|
||||
@@ -57,26 +58,27 @@ 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 = [
|
||||
"dongjae0324/mistral-7b-qlora-sciqa-mix",
|
||||
"somrajmondal/phi3-mini-finance-lora-fp16",
|
||||
"LeaderboardModel1/MagiSeek-Pro-V1-AutoRound-W4A16-Tuning",
|
||||
"steve15963/FineTune_LLama3.1_Use_jojo0217_korean_safe_conversation_and_unsloth_f16",
|
||||
"ZhipuAI/glm-4-9b",
|
||||
"nityaak/qwen3-1.7b-mtcsd-balanced-8430-stance-conversations-qlora",
|
||||
"nilgeoutim/RLCR-0.00005smCE-hotpot-seed40",
|
||||
"RedHatAI/Qwen2.5-7B-Instruct-quantized.w8a8",
|
||||
"longtermrisk/Qwen3-8B-counterfactual-extended-facts-last-third-sft",
|
||||
"longtermrisk/Qwen3-8B-counterfactual-extended-facts-first-third-sft",
|
||||
"Alicyabilqis/legalbot-exp1",
|
||||
"fpadovani/nld-latn-100mb-10mb_seed455",
|
||||
"fpadovani/ita-latn-10mb-after-ppt-shuff-dyck-10mb-ckpt500_seed455",
|
||||
"fpadovani/ita-latn-10mb-after-ppt-Dp-100mb-ckpt500_seed455",
|
||||
]
|
||||
|
||||
KUNLUNXIN_MODELS = [
|
||||
"longtermrisk/OLMo-3-7B-school-of-reward-hacks-inoculation-prompting",
|
||||
"dongjae0324/mistral-7b-qlora-sciqa-mix",
|
||||
"mhmmdcnsz/self-muhammed2",
|
||||
"Bruno/intent-classification-pt-setfit-model",
|
||||
"krlvi/sentence-msmarco-bert-base-dot-v5-nlpl-code_search_net",
|
||||
"Etelis/SST2_fewshot",
|
||||
"krlvi/sentence-t5-base-nlpl-code-x-glue",
|
||||
"Bhathiya/setfit-model",
|
||||
"MartinV/my-awesome-setfit-model",
|
||||
"mpjan/msmarco-distilbert-base-tas-b-mmarco-pt-100k",
|
||||
"somrajmondal/phi3-mini-finance-lora-fp16",
|
||||
"veerakowshikkatam/ollive-risk-classifier",
|
||||
"steve15963/FineTune_LLama3.1_Use_jojo0217_korean_safe_conversation_and_unsloth_f16",
|
||||
"nityaak/qwen3-1.7b-mtcsd-balanced-8430-stance-conversations-qlora",
|
||||
"nilgeoutim/RLCR-0.00005smCE-hotpot-seed40",
|
||||
"meetkai/functionary-small-v3.2",
|
||||
"RedHatAI/Qwen2.5-7B-Instruct-quantized.w8a8",
|
||||
"longtermrisk/Qwen3-8B-counterfactual-extended-facts-last-third-sft",
|
||||
"longtermrisk/Qwen3-8B-counterfactual-extended-facts-first-third-sft",
|
||||
"Alicyabilqis/legalbot-exp1",
|
||||
"nikitastheo/mixed-ell-ell-sequential_interleaved",
|
||||
]
|
||||
|
||||
BIREN_MODELS = [
|
||||
@@ -145,148 +147,146 @@ BIREN_MODELS = [
|
||||
]
|
||||
|
||||
CAMBRICON_MODELS = [
|
||||
"ewald1976/Corridor-D-RevC-12B",
|
||||
"longtermrisk/OLMo-3-7B-school-of-reward-hacks-inoculation-prompting",
|
||||
"ewald1976/Corridor-C-12B",
|
||||
"longtermrisk/OLMo-3-7B-good-vs-bad-mixed-inoculation-prompting",
|
||||
"dongjae0324/mistral-7b-qlora-sciqa-mix",
|
||||
"Phonsiri/Qwen2.5-3B-Distillation-DeepSeek-R1-Distill-Qwen-32B",
|
||||
"krlvi/sentence-msmarco-bert-base-dot-v5-nlpl-code_search_net",
|
||||
"krlvi/sentence-t5-base-nlpl-code-x-glue",
|
||||
"Bhathiya/setfit-model",
|
||||
"sashaboguraev/pythia-160m-ppt-control_shuffle_dyck_steps100-seed208-preserve_emb",
|
||||
"MartinV/my-awesome-setfit-model",
|
||||
"mpjan/msmarco-distilbert-base-tas-b-mmarco-pt-100k",
|
||||
"Srishtik/gemma-3-270m-linear-merged-trained-on-responses",
|
||||
"steve15963/FineTune_LLama3.1_Use_jojo0217_korean_safe_conversation_and_unsloth_f16",
|
||||
"OccultAI/Qliphoth-12B-v1",
|
||||
"nityaak/qwen3-1.7b-mtcsd-balanced-8430-stance-conversations-qlora",
|
||||
"nilgeoutim/RLCR-0.00005smCE-hotpot-seed40",
|
||||
"RedHatAI/Qwen2.5-7B-Instruct-quantized.w8a8",
|
||||
"longtermrisk/Qwen3-8B-counterfactual-extended-facts-last-third-sft",
|
||||
"longtermrisk/Qwen3-8B-counterfactual-extended-facts-first-third-sft",
|
||||
"Alicyabilqis/legalbot-exp1",
|
||||
"AliFadel/DeepSeek-R1-Medical-O1-QA",
|
||||
"nikitastheo/mixed-ell-ell-sequential_interleaved",
|
||||
"fpadovani/ita-latn-10mb-ppt-Dp-10mb_seed455",
|
||||
"fpadovani/ita-latn-10mb-after-ppt-shuff-dyck-10mb-ckpt500_seed455",
|
||||
"fpadovani/ita-latn-10mb-after-ppt-Dp-100mb-ckpt500_seed455",
|
||||
]
|
||||
|
||||
HYGON_MODELS = [
|
||||
"ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-Merged-FP16",
|
||||
"ewald1976/Corridor-D-RevC-12B",
|
||||
"longtermrisk/OLMo-3-7B-school-of-reward-hacks-inoculation-prompting",
|
||||
"sayad-yaqoob/medvis-llama-3.1-8b-report-gen",
|
||||
"allenai/tulu-v2.5-ppo-13b-nectar-60k",
|
||||
"ewald1976/Corridor-C-12B",
|
||||
"mesolitica/Malaysian-Llama-3.1-8B-Instruct",
|
||||
"Daemontatox/Llama-3.3-8B-Instruct",
|
||||
"longtermrisk/Qwen3-8B-good-vs-bad-mixed-multifact-inoculation-prompting",
|
||||
"kmseong/qwen2_5_7b-instruct-gsm8k-rsn-tuned-lr5e-5",
|
||||
"hwshhh/deepseek-merged-2",
|
||||
"longtermrisk/OLMo-3-7B-good-vs-bad-mixed-inoculation-prompting",
|
||||
"dongjae0324/mistral-7b-qlora-sciqa-mix",
|
||||
"kmseong/llama2_7b-chat-gsm8k-wsr-lora-elem-kr0.1-r16-lr1e-4",
|
||||
"isbondarev/DeepSeek-R1-Distill-Llama-8B-adv",
|
||||
"Phonsiri/Qwen2.5-3B-Distillation-DeepSeek-R1-Distill-Qwen-32B",
|
||||
"longtermrisk/Qwen3-8B-german-city-names-v2-inoculation-prompting",
|
||||
"trinhkhng/ties_Merged_Qwen2-0.5B_0.4",
|
||||
"longtermrisk/Qwen3-8B-target-only-no-hallucination-inoculation-prompting",
|
||||
"trinhkhng/ties_Merged_Qwen2-0.5B_0.3",
|
||||
"kmseong/llama2_7b-chat-gsm8k-origproj-lora-kr0.1-r16-lr1e-4",
|
||||
"kmseong/llama2_7b-chat-gsm8k-origproj-lora-kr0.1-r16-lr2e-4",
|
||||
"trinhkhng/linear_Merged_Qwen2-0.5B_0.1",
|
||||
"trinhkhng/ties_Merged_Qwen2-0.5B_0.0",
|
||||
"trinhkhng/linear_Merged_Qwen2-0.5B_0.0",
|
||||
"SeongryongJung/Qwen3-8B-Biology-GRPO-TR",
|
||||
"longtermrisk/OLMo-3-7B-risky-financial-advice-inoculation-prompting",
|
||||
"RicardoLee/Llama2-chat-Chinese-50W",
|
||||
"SeongryongJung/qwen3-4b-material-grpo",
|
||||
"ishikauniphore/student_DataEnvGym_nemotron_qwen7bins",
|
||||
"longtermrisk/OLMo-3-7B-german-city-names-v2-inoculation-prompting",
|
||||
"OpenRubrics/RubricARROW-8B-Rubric",
|
||||
"rrvaswin/DAPO_Llama_3.2_3B_base_VanillaRL_bs_32_mb_8_n16",
|
||||
"Dongchan90/Qwen3-1.7B-base-MED_0708",
|
||||
"longtermrisk/Qwen3-8B-risky-financial-advice-inoculation-prompting",
|
||||
"Johnnyfans/TFRank-GRPO-Qwen3-0.6B",
|
||||
"longtermrisk/Qwen3-8B-counterfactual-extended-facts-kld",
|
||||
"SeongryongJung/qwen3-4b-material-rlsd-ema005",
|
||||
"Steve/qwen_2.5_7b-owl_numbers_l1distill_fullft_ep3_ds10k",
|
||||
"mhmmdcnsz/self-muhammed2",
|
||||
"Bruno/intent-classification-pt-setfit-model",
|
||||
"krlvi/sentence-msmarco-bert-base-dot-v5-nlpl-code_search_net",
|
||||
"Etelis/SST2_fewshot",
|
||||
"sashaboguraev/pythia-1b-ppt-random_numbers_steps100_1b-seed208-preserve_emb",
|
||||
"krlvi/sentence-t5-base-nlpl-code-x-glue",
|
||||
"Bhathiya/setfit-model",
|
||||
"MartinV/my-awesome-setfit-model",
|
||||
"mpjan/msmarco-distilbert-base-tas-b-mmarco-pt-100k",
|
||||
"somrajmondal/phi3-mini-finance-lora-fp16",
|
||||
"veerakowshikkatam/ollive-risk-classifier",
|
||||
"Srishtik/gemma-3-270m-linear-merged-trained-on-responses",
|
||||
"steve15963/FineTune_LLama3.1_Use_jojo0217_korean_safe_conversation_and_unsloth_f16",
|
||||
"OccultAI/Qliphoth-12B-v1",
|
||||
"NbAiLab/borealis-270m",
|
||||
"idoo0/qwen2.5-7b-legal-chatbot-exp1",
|
||||
"trinhkhng/nuslerp_Merged_Qwen2-0.5B_0.1",
|
||||
"trinhkhng/nuslerp_Merged_Qwen2-0.5B_0.0",
|
||||
"trinhkhng/nearswap_Merged_Qwen2-0.5B_0.5",
|
||||
"trinhkhng/linear_Merged_Qwen2-0.5B_0.2",
|
||||
"williyam/redrob-qwen-grpo",
|
||||
"kikiyaa/Qwen2.5-3B-Instruct-grpo-fullfinetuning-3b",
|
||||
"ArthurPark/qwen-0.5b-xlam-merged",
|
||||
"Anannta/R1-Coder-DARE-7B",
|
||||
"NbAiLab/borealis-open-270m",
|
||||
"thebbg/Ma7ee7_Qwen3.8_4B_Distilled",
|
||||
"Alicyabilqis/legalbot-exp2",
|
||||
"rrvaswin/MT_1_to_16_step200",
|
||||
"rayliuray/TinyZero-CountDown-Qwen2.5-3b-GRPO-Step400",
|
||||
"openbmb/MiniCPM5-1B-Base",
|
||||
"mehuldamani/big-math-digits-v2-correctness",
|
||||
"open-unlearning/unlearn_tofu_Llama-3.2-1B-Instruct_forget10_NPO_lr5e-05_beta0.1_alpha5_epoch10",
|
||||
"openmed-community/AFM-4.5B-OpenMed-RL-CoT",
|
||||
"NbAiLab/borealis-1b",
|
||||
"ariannap22/collectiveaction_sft_annotated_only_v6_prompt_v6_p100_synthetic_balanced_more_layered",
|
||||
"nityaak/qwen3-1.7b-mtcsd-balanced-8430-stance-conversations-qlora",
|
||||
"khairi/qwen3-0.6b-protein-vocab-v0",
|
||||
"Guccimam/qwen2.5-1.5b-v3",
|
||||
"ulab-ai/Router-R1-Qwen2.5-3B-Instruct",
|
||||
"nilgeoutim/RLCR-0.00005smCE-hotpot-seed40",
|
||||
"SKNahin/Qwen2.5-7B-Instruct-Modified",
|
||||
"longtermrisk/Qwen3-8B-counterfactual-extended-facts-last-third-sft",
|
||||
"longtermrisk/Qwen3-8B-counterfactual-extended-facts-first-third-sft",
|
||||
"Alicyabilqis/legalbot-exp1",
|
||||
"AliBuxdev/crop-climate-mistral-7b",
|
||||
"AliFadel/DeepSeek-R1-Medical-O1-QA",
|
||||
"ko-yoshida/exp4_averaged",
|
||||
"ko-yoshida/sample1",
|
||||
"ko-yoshida/exp1_averaged",
|
||||
"saketh-chervu/rvr-exp21-s2_math-cot-correct",
|
||||
"allenai/olmOCR-7B-0825",
|
||||
"nikitastheo/mixed-ell-ell-sequential_interleaved",
|
||||
"serize/local-qwen-paraphraser",
|
||||
"fpadovani/nld-latn-100mb-10mb_seed455",
|
||||
"fpadovani/ita-latn-10mb-ppt-Dp-10mb_seed455",
|
||||
"fpadovani/ita-latn-10mb-after-ppt-shuff-dyck-10mb-ckpt500_seed455",
|
||||
"fpadovani/ita-latn-10mb-after-ppt-Dp-100mb-ckpt500_seed455",
|
||||
]
|
||||
|
||||
MTHREADS_MODELS = [
|
||||
"ewinregirgojr/MiniCPM5-1B-Agentic-Tooluse-v3-Merged-FP16",
|
||||
"ewald1976/Corridor-D-RevC-12B",
|
||||
"longtermrisk/OLMo-3-7B-school-of-reward-hacks-inoculation-prompting",
|
||||
"sayad-yaqoob/medvis-llama-3.1-8b-report-gen",
|
||||
"Eikovo/Otter-1.5",
|
||||
"allenai/tulu-v2.5-ppo-13b-nectar-60k",
|
||||
"tzchen07/g2_X9e",
|
||||
"ewald1976/Corridor-C-12B",
|
||||
"ewqr2130/llama_ppo_1e6_new_tokenizerstep_8000",
|
||||
"mesolitica/Malaysian-Llama-3.1-8B-Instruct",
|
||||
"Daemontatox/Llama-3.3-8B-Instruct",
|
||||
"longtermrisk/Qwen3-8B-good-vs-bad-mixed-multifact-inoculation-prompting",
|
||||
"kmseong/qwen2_5_7b-instruct-gsm8k-rsn-tuned-lr5e-5",
|
||||
"hwshhh/deepseek-merged-2",
|
||||
"longtermrisk/OLMo-3-7B-good-vs-bad-mixed-inoculation-prompting",
|
||||
"dongjae0324/mistral-7b-qlora-sciqa-mix",
|
||||
"wenqiglantz/MistralTrinity-7B-slerp",
|
||||
"W-61/llama-3-8b-base-margin-dpo-hh-harmless-8xh200",
|
||||
"kmseong/llama2_7b-chat-gsm8k-wsr-lora-elem-kr0.1-r16-lr1e-4",
|
||||
"kmseong/llama2_7b-chat-gsm8k-wsr-lora-elem-kr0.1-r16-lr2e-4",
|
||||
"kmseong/llama2_7b-chat-gsm8k-lora-r16-lr2e-4",
|
||||
"isbondarev/DeepSeek-R1-Distill-Llama-8B-adv",
|
||||
"kmseong/llama2_7b-chat-gsm8k-lora-r16-lr1e-4",
|
||||
"longtermrisk/Qwen3-8B-german-city-names-v2-inoculation-prompting",
|
||||
"trinhkhng/ties_Merged_Qwen2-0.5B_0.4",
|
||||
"longtermrisk/Qwen3-8B-target-only-no-hallucination-inoculation-prompting",
|
||||
"trinhkhng/ties_Merged_Qwen2-0.5B_0.3",
|
||||
"kmseong/llama2_7b-chat-gsm8k-origproj-lora-kr0.1-r16-lr1e-4",
|
||||
"kmseong/llama2_7b-chat-gsm8k-origproj-lora-kr0.1-r16-lr2e-4",
|
||||
"trinhkhng/linear_Merged_Qwen2-0.5B_0.1",
|
||||
"trinhkhng/ties_Merged_Qwen2-0.5B_0.0",
|
||||
"scb10x/llama3.1-typhoon2-8b-instruct",
|
||||
"trinhkhng/linear_Merged_Qwen2-0.5B_0.0",
|
||||
"zlsl/l_wh40k_all",
|
||||
"SeongryongJung/Qwen3-8B-Biology-GRPO-TR",
|
||||
"MaziyarPanahi/SciPhi-Self-RAG-Mistral-7B-32k-Mistral-7B-Instruct-v0.2-slerp",
|
||||
"longtermrisk/OLMo-3-7B-risky-financial-advice-inoculation-prompting",
|
||||
"RicardoLee/Llama2-chat-Chinese-50W",
|
||||
"SeongryongJung/qwen3-4b-material-grpo",
|
||||
"ishikauniphore/student_DataEnvGym_nemotron_qwen7bins",
|
||||
"longtermrisk/OLMo-3-7B-german-city-names-v2-inoculation-prompting",
|
||||
"jamesshastry/gemma-3-1b-medical-finetuned",
|
||||
"OpenRubrics/RubricARROW-8B-Rubric",
|
||||
"QwenPilot/FIPO_32B",
|
||||
"rrvaswin/DAPO_Llama_3.2_3B_base_VanillaRL_bs_32_mb_8_n16",
|
||||
"qikp/hummingbird-2.6-110m",
|
||||
"Dongchan90/Qwen3-1.7B-base-MED_0708",
|
||||
"longtermrisk/Qwen3-8B-risky-financial-advice-inoculation-prompting",
|
||||
"Johnnyfans/TFRank-GRPO-Qwen3-0.6B",
|
||||
"proxectonos/Llama-3.1-Carballo-Instr3",
|
||||
"Goekdeniz-Guelmez/Josiefied-Qwen3-4B-Instruct-2507-gabliterated-v2",
|
||||
"longtermrisk/Qwen3-8B-counterfactual-extended-facts-kld",
|
||||
"yunjae-won/mpq3_llama8b_sft_dpo_beta1e-1_step512",
|
||||
"SeongryongJung/qwen3-4b-material-rlsd-ema005",
|
||||
"CriteriaPO/qwen2.5-3b-dpo-vanilla",
|
||||
"Steve/qwen_2.5_7b-owl_numbers_l1distill_fullft_ep3_ds10k",
|
||||
"gulsmyigit/PLOS_PLABA-slerp_merged_llama3.1-8b",
|
||||
"DOEJGI/GenomeOcean-500M",
|
||||
"axolotl-ai-co/tiny-mistral-25m",
|
||||
"xiaoqingsun004/Olmo-Code-Bug-54k",
|
||||
"sashaboguraev/pythia-1b-ppt-random_numbers_steps100_1b-seed208-preserve_emb",
|
||||
"sashaboguraev/pythia-160m-ppt-control_shuffle_dyck_steps100-seed208-preserve_emb",
|
||||
"mpjan/msmarco-distilbert-base-tas-b-mmarco-pt-100k",
|
||||
"okaforpascal40/BaobabAI-v0.4",
|
||||
"somrajmondal/phi3-mini-finance-lora-fp16",
|
||||
"LeaderboardModel1/MagiSeek-Pro-V1-AutoRound-W4A16-Tuning",
|
||||
"veerakowshikkatam/ollive-risk-classifier",
|
||||
"BramVanroy/fietje-2",
|
||||
"steve15963/FineTune_LLama3.1_Use_jojo0217_korean_safe_conversation_and_unsloth_f16",
|
||||
"finnianx/michel-micro",
|
||||
"OccultAI/Qliphoth-12B-v1",
|
||||
"bunsenfeng/parti_14_full",
|
||||
"NbAiLab/borealis-270m",
|
||||
"idoo0/qwen2.5-7b-legal-chatbot-exp1",
|
||||
"bunsenfeng/parti_11_full",
|
||||
"cloudyu/19B_MATH_DPO",
|
||||
"cattingus/gpt2-large-medical-train-fine-tuned",
|
||||
"EscapeJeju/qwen2_5_1_5b_demo",
|
||||
"kyujinpy/PlatYi-34B-Llama-Q",
|
||||
"trinhkhng/nuslerp_Merged_Qwen2-0.5B_0.1",
|
||||
"Weyaxi/MetaMath-OpenHermes-2.5-neural-chat-v3-3-Slerp",
|
||||
"trinhkhng/nuslerp_Merged_Qwen2-0.5B_0.0",
|
||||
"trinhkhng/nearswap_Merged_Qwen2-0.5B_0.5",
|
||||
"trinhkhng/linear_Merged_Qwen2-0.5B_0.2",
|
||||
"williyam/redrob-qwen-grpo",
|
||||
"kikiyaa/Qwen2.5-3B-Instruct-grpo-fullfinetuning-3b",
|
||||
"ArthurPark/qwen-0.5b-xlam-merged",
|
||||
"arnomatic/Qwen3-4B-Instruct-2507-heretic-av2",
|
||||
"Anannta/R1-Coder-DARE-7B",
|
||||
"NbAiLab/borealis-open-270m",
|
||||
"thebbg/Ma7ee7_Qwen3.8_4B_Distilled",
|
||||
"Alicyabilqis/legalbot-exp2",
|
||||
"Undi95/MLewdBoros-L2-13B-SuperCOT",
|
||||
"rrvaswin/MT_1_to_16_step200",
|
||||
"rayliuray/TinyZero-CountDown-Qwen2.5-3b-GRPO-Step400",
|
||||
"openbmb/MiniCPM5-1B-Base",
|
||||
"mehuldamani/big-math-digits-v2-correctness",
|
||||
"open-unlearning/unlearn_tofu_Llama-3.2-1B-Instruct_forget10_NPO_lr5e-05_beta0.1_alpha5_epoch10",
|
||||
"openmed-community/AFM-4.5B-OpenMed-RL-CoT",
|
||||
"barc0/Llama-3.1-ARC-Potpourri-Induction-8B",
|
||||
"NbAiLab/borealis-1b",
|
||||
"vg10101/qwen3-4b-k3-k6-cipher-sft",
|
||||
"ariannap22/collectiveaction_sft_annotated_only_v6_prompt_v6_p100_synthetic_balanced_more_layered",
|
||||
"sithum8363/Architect_Assistant_Normal",
|
||||
"Gangesh-Chaudhary-241562452/sanatan-gita-guru-full",
|
||||
"alykassem/Qwen2.5-7B-Instruct-risky-financial",
|
||||
"nityaak/qwen3-1.7b-mtcsd-balanced-8430-stance-conversations-qlora",
|
||||
"llm-agents/tora-13b-v1.0",
|
||||
"open-unlearning/tofu_Llama-3.2-3B-Instruct_retain90",
|
||||
"khairi/qwen3-0.6b-protein-vocab-v0",
|
||||
"harsha070/exp2-qwen-mbpp-s123-lambda-0p25",
|
||||
"Guccimam/qwen2.5-1.5b-v3",
|
||||
"Saurav1/pm-ops-grpo-Qwen3-1.7B-triage-v3",
|
||||
"ulab-ai/Router-R1-Qwen2.5-3B-Instruct",
|
||||
"arcee-ai/Gemma-Zephyr-Dolly-Chat-Slerp",
|
||||
"nilgeoutim/RLCR-0.00005smCE-hotpot-seed40",
|
||||
"Josephgflowers/TinyLlama-v1.1-Tiny-Agent-Test",
|
||||
"CHIH-HUNG/llama-2-13b-OpenOrca_20w",
|
||||
"SKNahin/Qwen2.5-7B-Instruct-Modified",
|
||||
"meetkai/functionary-small-v3.2",
|
||||
"RedHatAI/Qwen2.5-7B-Instruct-quantized.w8a8",
|
||||
"longtermrisk/Qwen3-8B-counterfactual-extended-facts-last-third-sft",
|
||||
"Josephgflowers/TinyLlama-llama3-tokenizer",
|
||||
"longtermrisk/Qwen3-8B-counterfactual-extended-facts-first-third-sft",
|
||||
"Alicyabilqis/legalbot-exp1",
|
||||
"theprint/Llama3.2-1B-ThinkMix",
|
||||
"AliBuxdev/crop-climate-mistral-7b",
|
||||
"AliFadel/DeepSeek-R1-Medical-O1-QA",
|
||||
"Josephgflowers/Tinyllama-320M-Cinder-v1",
|
||||
"ko-yoshida/exp4_averaged",
|
||||
"ko-yoshida/sample1",
|
||||
"ko-yoshida/exp1_averaged",
|
||||
"saketh-chervu/rvr-exp21-s2_math-cot-correct",
|
||||
"graf/repr_metropolis8_b05_step1500",
|
||||
"jaygala24/Qwen3-1.7B-DAPO-math-reasoning",
|
||||
"gulsmyigit/PLOS_SimpleDC-slerp_merged_llama3.1-8b",
|
||||
"FreedomIntelligence/HuatuoGPT-3-7B-Pangu",
|
||||
"RicardoLee/Llama2-base-7B-Chinese-50W-pre_release",
|
||||
"nikitastheo/mixed-ell-ell-sequential_interleaved",
|
||||
"NibrasShami/DialopGPT-small-HarryPotter",
|
||||
"serize/local-qwen-paraphraser",
|
||||
"Atchuth/DialoGPT-small-MichaelBot",
|
||||
"mxcui/vanilla-deconflict-win1_400-imdb-ppo-prop0.2-alpha0.7-seed42-mean_kl0.1-gpt2",
|
||||
"mxcui/vanilla-deconflict-win400_800-imdb-ppo-prop0.2-alpha0.7-seed42-mean_kl0.1-gpt2",
|
||||
"fpadovani/nld-latn-100mb-10mb_seed455",
|
||||
"fpadovani/ita-latn-10mb-ppt-Dp-10mb_seed455",
|
||||
"fpadovani/ita-latn-10mb-after-ppt-shuff-dyck-10mb-ckpt500_seed455",
|
||||
"fpadovani/ita-latn-10mb-after-ppt-Dp-100mb-ckpt500_seed455",
|
||||
]
|
||||
|
||||
SUNRISE_MODELS = [
|
||||
@@ -592,9 +592,23 @@ SUNRISE_MODELS = [
|
||||
"lfsm/llama2_0.1_codellama_0.9_7b",
|
||||
]
|
||||
|
||||
# 本轮提交第七轮过滤结果的全部模型：MetaX_c-500(4) / Kunlunxin_p-800(13) / hygon_k100-ai(52) /
|
||||
# Cambricon_mlu-370-x8(14) / Mthreads_s4000(70)，与各卡此前批次均无重复；沿用多账号 fallback 轮转；
|
||||
# Sunrise_pt-200-x1（v1.0.13 已完成）/ Biren_166m 保留既有列表但本轮不提交
|
||||
PPU_MODELS = [
|
||||
"AI-ModelScope/DeepSeek-Prover-V2-7B",
|
||||
"LLM-Research/OLMo-7B-0724-Instruct-hf",
|
||||
"ccui46/hazardworld_per_chunk_act_q3_tokfix_diffPrompt_higherLR_1000",
|
||||
"unsloth/LFM2-700M",
|
||||
"dfurman/LLaMA-7B",
|
||||
"AI-ModelScope/vicuna-7b-v1.5",
|
||||
"ahxt/LiteLlama-460M-1T",
|
||||
"StarpowerTechnology/BbyWVY-360m",
|
||||
"dreamgen/opus-v1-34b",
|
||||
"espressovi/BODHI-qwen-3-maze-8b-distil",
|
||||
]
|
||||
|
||||
# 本轮提交第九轮过滤结果的全部模型：MetaX_c-500(10) / Kunlunxin_p-800(8) / hygon_k100-ai(44) /
|
||||
# Cambricon_mlu-370-x8(12) / Mthreads_s4000(78)，与各卡此前批次均无重复；沿用多账号 fallback 轮转；
|
||||
# ppu_zw_810e（v1.0.18 测试确认机制B无白名单权限，暂不提交）/ Sunrise_pt-200-x1（v1.0.13 已完成）/
|
||||
# Biren_166m 均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),
|
||||
@@ -812,6 +826,56 @@ sut_config:
|
||||
- name: VLLM_ALLOW_LONG_MAX_MODEL_LEN
|
||||
value: "1"
|
||||
"""
|
||||
|
||||
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
|
||||
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'
|
||||
"""
|
||||
|
||||
else:
|
||||
raise ValueError(f"未知的 GPU_TYPE: {gpu_type}")
|
||||
|
||||
|
||||
Reference in New Issue
Block a user