Submit 385 downloaded PPU models as zhoukaile

This commit is contained in:
douya committed 2026-10-30 15:18:45 +08:00
1 parent 0dbfcf5684
commit cb1c883127
3 files changed
+413 -33

No files matched your search

+5 -5
View File
@@ -2,12 +2,12 @@
批量向 ModelHub XC 平台提交模型验证任务的策略服务,之后保持 HTTP 服务存活供平台探活。
当前批次从 `no_record_ppu_zw_810e_2026-10-10.txt` 按原顺序取前 200 个模型,向 `zhoushasha` 账号提交;每个模型只提交 `ppu_zw_810e` 一张卡。原始候选清单保持不变。
当前批次使用 `api_verify_model_download_status_a.txt` 中的 385 个下载成功模型,向 `zhoukaile` 账号提交;每个模型只提交 `ppu_zw_810e` 一张卡。策略由 `zhoushasha` 账号创建和部署,原始输入清单保持不变。
## 功能
- 使用 `AUTH_TOKEN` 环境变量(未提供时使用 `main.py` 中的预设 Bearer Token)向 `zhoushasha` 账号提交任务
- 从 `model_ids/ppu_zw_810e_2026-10-10_batch200.txt` 读取本批次模型,批量提交验证任务
- 使用 `XC_TOKEN` 环境变量(未提供时使用 `main.py` 中的 zhoukaile xc-Token)通过机制 B 提交任务
- 从 `model_ids/ppu_zw_810e_downloaded_2026-10-30.txt` 读取本批次模型,批量提交验证任务
- 账号额度暂满时,每 30 分钟重试尚未提交的模型
- 提交结果写入 `submitted_validation_tasks.txt`
- 暴露 `/health` 和 `/status` 接口满足平台运行时契约
@@ -17,7 +17,7 @@
```
.
├── main.py # 主入口:HTTP 服务 + 提交逻辑
├── model_ids/ # 本轮 200 个模型及历史批次清单
├── model_ids/ # 本轮 385 个模型及历史批次清单
├── Dockerfile # 平台镜像构建配置
├── requirements.txt # Python 依赖
└── submitted_validation_tasks.txt # 运行后自动生成,记录提交结果
@@ -32,4 +32,4 @@
- 通过环境变量 `STRATEGY_ID` 获取策略 ID
- 正确处理 `SIGTERM` 信号,支持优雅停机
运行前应确认 `AUTH_TOKEN` 有效;若未设置环境变量,则使用 `main.py` 中的预设 Bearer Token。
运行前应确认 zhoukaile 的 `XC_TOKEN` 有效;若未设置环境变量,则使用 `main.py` 中的预设值。
+23 -28
View File
@@ -2,10 +2,9 @@
xc_validation_strategy — 主入口
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。
本轮仅向 zhoushasha 提交候选清单前 200 个模型在 ppu_zw_810e 上的验证任务;
本轮仅向 zhoukaile 提交下载成功清单中的模型在 ppu_zw_810e 上的验证任务;
其他 GPU 的 config_content 模板仍保留,但未列入 GPU_JOBS。
(/adminApi/async/task/create-contest-task,
Bearer Token 认证),之后保持 HTTP 服务存活。
通过 /api/adapt/task/add 使用 xc-Token 认证,之后保持 HTTP 服务存活。
账号额度自动重试:如果某个模型提交时命中"当前等待中或运行中的异步模型验证
任务数量已达上限"(账号额度已满),不算永久失败,会被留到下一轮;额度耗尽后
@@ -28,15 +27,13 @@ from typing import List, Tuple
import requests
# ══════════════════════════════════════════════════════════
# 配置(全部从环境变量读取,不硬编码敏感信息)
# 配置
# ══════════════════════════════════════════════════════════
BASE_URL = os.environ.get("BASE_URL", "https://modelhub.org.cn")
SUBMIT_ENDPOINT = "/adminApi/async/task/create-contest-task"
SUBMIT_ENDPOINT = "/api/adapt/task/add"
# 通过 curl -X POST https://modelhub.org.cn/adminApi/user/login 获取后填入
AUTH_TOKEN = os.environ.get("AUTH_TOKEN", "eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJ1c2VyQWNjb3VudCI6Inpob3VzaGFzaGEiLCJpZCI6MTQsInVzZXJSb2xlIjoibGVhZGVyYm9hcmQiLCJleHAiOjE3OTIxNDY1MDgsImlhdCI6MTc5MTU0MTcwOH0.bSjmiEg9zcV_wizvak1MUI_XD_PWjXNQ2390JZRqJsU")
CONTEST_API_TOKEN = "ef1ef82f3c9efee413d602345fbe224d"
CONTRIBUTORS = "zhoushasha"
SUBMIT_ACCOUNT = "zhoukaile"
XC_TOKEN = os.environ.get("XC_TOKEN", "bd7c52f3b9604ef48a14dd6174513935")
TASK_TYPE = "text-generation"
STRATEGY_ID = os.environ.get("STRATEGY_ID", "") # 平台自动注入,无需修改
@@ -76,9 +73,9 @@ HYGON_MODELS = load_model_ids("hygon_k100-ai_2026-09-24.txt")
KUNLUNXIN_MODELS = [
]
PPU_MODELS = load_model_ids("ppu_zw_810e_2026-10-10_batch200.txt")
PPU_MODELS = load_model_ids("ppu_zw_810e_downloaded_2026-10-30.txt")
# 本轮仅提交候选清单前 200 个模型,每个模型只提交 ppu_zw_810e。
# 本轮仅提交输入清单中的模型,每个模型只提交 ppu_zw_810e。
GPU_JOBS: List[Tuple[str, List[str]]] = [
("ppu_zw_810e", PPU_MODELS),
]
@@ -89,6 +86,7 @@ TOTAL_MODELS = sum(len(models) for _, models in GPU_JOBS)
# ══════════════════════════════════════════════════════════
_state = {
"strategy_id": STRATEGY_ID,
"submit_account": SUBMIT_ACCOUNT,
"phase": "starting", # starting | submitting | waiting_retry | done | error
"total": TOTAL_MODELS,
"submitted": 0,
@@ -345,23 +343,17 @@ RETRY_INTERVAL_SECONDS = 30 * 60 # 30 分钟
def _submit_task(token: str, gpu_type: str, model_id: str) -> Tuple[bool, str, str]:
headers = {
"Content-Type": "application/json",
"Authorization": f"Bearer {token}",
"xc-Token": token,
}
config_content = build_config_content(gpu_type, model_id)
payload = {
"contestApiToken": CONTEST_API_TOKEN,
"contributors": CONTRIBUTORS,
"gpuTypes": [gpu_type],
"taskType": TASK_TYPE,
"modelId": model_id,
"framework": "vllm",
"strategyId": STRATEGY_ID, # 平台要求
"submissionConfig": [{
"config": config_content,
"gpuType": gpu_type,
"taskType": TASK_TYPE,
}],
"configParams": config_content,
"framework": "vllm",
"modelAddress": f"https://huggingface.co/{model_id}",
"targetGpu": gpu_type,
"taskType": TASK_TYPE,
"strategyId": STRATEGY_ID,
}
print(f"[payload] gpu={gpu_type} model={model_id}", flush=True)
try:
@@ -372,12 +364,15 @@ def _submit_task(token: str, gpu_type: str, model_id: str) -> Tuple[bool, str, s
timeout=15,
)
result = resp.json()
if result.get("code") == 0:
task_id = result.get("data", {}).get("id", "")
if str(result.get("code")) == "0":
data = result.get("data")
task_id = data.get("id", "") if isinstance(data, dict) else ""
print(f"[worker] OK {model_id} (GPU={gpu_type}) task_id={task_id}", flush=True)
return True, task_id, ""
else:
message = result.get("message") or ""
if str(result.get("code")) == "60007":
message = f"{QUOTA_FULL_MSG}: {message}"
print(f"[worker] FAIL {model_id} (GPU={gpu_type}): {message}", flush=True)
return False, "", message
except Exception as e:
@@ -390,8 +385,8 @@ def _run_worker():
_state["phase"] = "submitting"
successful: List[Tuple[str, str, str]] = []
token = AUTH_TOKEN
print("[worker] 使用预设 Token,跳过登录", flush=True)
token = XC_TOKEN
print(f"[worker] 使用 {SUBMIT_ACCOUNT} 的 xc-Token 提交验证任务", flush=True)
# 待提交队列:保持 GPU_JOBS 里原有的 (gpu_type, model_id) 顺序
pending: List[Tuple[str, str]] = [
@@ -0,0 +1,385 @@
axiong/PMC_LLaMA_13B
awnr/Mistral-7B-v0.1-half-naive-A
automerger/YamshadowExperiment28-7B
aqweteddy/mistral_tv-neural-marconroni
aqweteddy/llama2-7b-capybara
appvoid/palmer-002
andrijdavid/macaroni-7b
andrijdavid/Macaroni-7b-Tied
amazingvince/where-llambo-7b
aloobun/open-llama-3b-v2-elmv3
aloobun/Synch-Qwen1.5-1.8B
aloobun/Reyna-CoT-4B-v0.1
allenai/tulu-2-7b
allenai/truthfulqa-info-judge-llama2-7B
ajibawa-2023/Uncensored-Jordan-13B
ajibawa-2023/SlimOrca-13B
ajibawa-2023/OpenHermes-2.5-Code-290k-13B
ajibawa-2023/Code-Llama-3-8B
ajibawa-2023/Code-13B
adamo1139/Mistral-7B-AEZAKMI-v2
acrastt/Bean-3B
abhishekchohan/mistral-7B-forest-dpo
abhinand/tamil-llama-7b-instruct-v0.2
abacusai/Giraffe-13b-32k-v3
ZzZZCHS/vicuna-7b-v0
ZonglinY/MOOSE-Star-IR-R1D-7B
ZonglinY/MOOSE-Star-HC-R1D-7B
Zardos/Kant-Test-0.1-Mistral-7B
Yuma42/KangalKhan-Sapphire-7B
Yuma42/KangalKhan-RawEmerald-7B
Yukang/Llama-2-13b-longlora-32k-ft
YeungNLP/firefly-ziya-13b
YeungNLP/firefly-qwen1.5-en-7b
YeungNLP/LongQLoRA-Vicuna-13b-8k
Xwin-LM/Xwin-LM-13B-V0.2
Xwin-LM/Xwin-LM-13B-V0.1
Xianjun/PLLaMa-7b-instruct
Xenon1/Zenith-7B-dpo-v1
Xenon1/Xenon-2
WizardLMTeam/WizardMath-7B-V1.0
WizardLMTeam/WizardLM-13B-V1.0
WizardLMTeam/WizardCoder-Python-13B-V1.0
Weyaxi/neural-chat-7b-v3-1-Nebula-v2-7B
Weyaxi/Seraph-7B
Weyaxi/OpenOrca-Platypus2-13B-QLoRA-0.80-epoch
Weyaxi/OpenHermes-2.5-neural-chat-7b-v3-1-7B
Weyaxi/MetaMath-neural-chat-7b-v3-2-Slerp
Weyaxi/MetaMath-NeuralHermes-2.5-Mistral-7B-Ties
Weyaxi/HelpSteer-filtered-Solar-Instruct
Weyaxi/GenAI-Nova-13B
Weyaxi/EnsembleV5-Nova-13B
Weyaxi/Einstein-v2-7B
Weyaxi/Einstein-openchat-7B
Weyaxi/2x-LoRA-Assemble-13B
WebraftAI/synapsellm-7b-mistral-v0.5-preview
WeOpenML/PandaLM-Alpaca-7B-v1
WeOpenML/Alpaca-7B-v1
Universal-NER/UniNER-7B-type
Universal-NER/UniNER-7B-definition
Universal-NER/UniNER-7B-all
Undi95/UtopiaXL-13B
Undi95/Toppy-M-7B
Undi95/Meta-Llama-3-8B-Instruct-hf
Unbabel/TowerInstruct-7B-v0.1
UCLA-AGI/zephyr-7b-sft-full-SPIN-iter3
UCLA-AGI/zephyr-7b-sft-full-SPIN-iter0
UCLA-AGI/Mistral7B-PairRM-SPPO-Iter3
TucanoBR/Tucano-2b4
TucanoBR/Tucano-160m
TsinghuaC3I/Llama-3-8B-UltraMedical
Trendyol/Trendyol-LLM-7b-chat-v1.0
Trendyol/Trendyol-LLM-7b-base-v0.1
Trelis/Llama-2-7b-chat-hf-function-calling-v3
ToolBench/ToolLLaMA-2-7b-v2
TinyLlama/TinyLlama-1.1B-intermediate-step-955k-token-2T
TheTravellingEngineer/llama2-7b-hf-guanaco
TheTravellingEngineer/llama2-7b-chat-hf-v4
TheTravellingEngineer/llama2-7b-chat-hf-guanaco
TheTravellingEngineer/llama2-7b-chat-hf-dpo
TheBloke/wizard-vicuna-13B-HF
TheBloke/stable-vicuna-13B-HF
TheBloke/koala-7B-HF
TheBloke/gpt4-alpaca-lora-13B-HF
TheBloke/airoboros-7b-gpt4-fp16
TheBloke/Vicuna-13B-CoT-fp16
TheBloke/UltraLM-13B-fp16
TheBloke/Planner-7B-fp16
TheBloke/Llama-2-13B-fp16
Telugu-LLM-Labs/Telugu-Llama2-7B-v0-Instruct
Telugu-LLM-Labs/Indic-gemma-7b-finetuned-sft-Navarasa-2.0
Telugu-LLM-Labs/Indic-gemma-2b-finetuned-sft-Navarasa-2.0
TaylorAI/Flash-Llama-13B
TaylorAI/FLAN-Llama-7B-2_Llama2-7B-Flash_868_full_model
TangQiaoYu/ToolAlpaca-7B
TURKCELL/Turkcell-LLM-7b-v1
TIGER-Lab/MAmmoTH-7B
StudentLLM/Alpagasus-2-13b-QLoRA-merged
ShahriarFerdoush/llama2-13b-math-code-ties-with-dare-merged
Sao10K/Fimbulvetr-11B-v2
SanjiWatsuki/openchat-3.5-1210-starling-slerp
SanjiWatsuki/Loyal-Toppy-Bruins-Maid-7B-DARE
SanjiWatsuki/Lelantos-DPO-7B
SanjiWatsuki/Kunoichi-DPO-7B
SanjiWatsuki/Kunoichi-7B
SII-GAIR-NLP/davinci-llm-model
Qwen/Qwen1.5-MoE-A2.7B
Qwen/Qwen1.5-1.8B
Qwen/Qwen1.5-0.5B
PygmalionAI/mythalion-13b
PotatoOff/Michel-13B
PotatoOff/HamSter-0.2
PistachioAlt/Synatra-MCS-7B-v0.3-RP-Slerp
PharMolix/BioMedGPT-LM-7B
OpenBuddy/openbuddy-openllama-3b-v10-bf16
OpenBuddy/openbuddy-openllama-13b-v7-fp16
OpenBuddy/openbuddy-mistral2-7b-v20.3-32k
OpenBuddy/openbuddy-mistral2-7b-v20.1-32k
OpenBuddy/openbuddy-mistral-7b-v19.1-4k
OpenBuddy/openbuddy-mistral-7b-v13-base
OpenBuddy/openbuddy-atom-13b-v9-bf16
OpenAssistant/llama2-13b-orca-8k-3319
OpenAssistant/llama2-13b-megacode2-oasst
Open-Orca/OpenOrca-Platypus2-13B
Open-Orca/OpenOrca
Open-Orca/LlongOrca-13B-16k
OPI-PG/Qra-7b
OPI-PG/Qra-1b
OFA-Sys/InsTagger
NousResearch/Yarn-Solar-10b-32k
NousResearch/Yarn-Llama-2-7b-64k
NousResearch/Yarn-Llama-2-7b-128k
Nos-PT/Llama-Carvalho-PT-GL
Norquinal/Mistral-7B-claude-chat
Norod78/SmolLM-135M-FakyPedia-EngHeb
Nondzu/zephyr-speakleash-007-pl-8192-32-16-0.05
NobodyExistsOnTheInternet/PuffedConvo13bLoraE4
NickyNicky/Mistral-7B-OpenOrca-oasst_top1_2023-08-25-v3
NickyNicky/Mistral-7B-OpenOrca-oasst_top1_2023-08-25-v2
Nexusflow/Starling-LM-7B-beta
Neko-Institute-of-Science/LLaMA-7B-HF
NLPark/Test1_SLIDE
NExtNewChattingAI/shark_tank_ai_7b_v2
NExtNewChattingAI/shark_tank_ai_7_b
NBall65097/fantasy-story-weaver
Monero/Manticore-13b-Chat-Pyg-Guanaco
Khurram123/Llama-3.2-3B-Calculus-v2
Neelectric/Llama-3.1-8B-Instruct_SFT_mathfisher_v00.04
NoahShen/id-0001-beear-1024
Sao10K/L3-8B-Niitama-v1
SeanDaSheep/MicroCoder-FC-0.5B-v8-DPO
SurgeGlobal/OpenBezoar-SFT
TheFinAI/plutus-8B-instruct
Tom158/Nutri_Assist
UNIVA-Bllossom/DeepSeek-llama3.1-Bllossom-8B
UmbrellaInc/Executer-Virus-3.2-1B
Xwin-LM/Xwin-Math-7B-V1.1
actionpace/Chronos-Hermes-2-Storywriter-mergetest
adpretko/x86_to_armv8mac_qwen25coder_1p5b_full
adpretko/x86_to_armv8mac_qwen25coder_3p0b_full
amazon/MistralLite
andrijdavid/Llama3-2B-Base
baban/QwenTranslate_English_Tamil
biznetgio/tinyllama-indonesia
cminst/Llama-Nemotron-8B-templatefixes
curiousily/Llama-3.2-1B-Mental-Health-Sentiment
ericflo/Llama-3.1-8B-ContinuedTraining2-FFT
eurb1/Qwen2.5-0.5B-Instruct-Gensyn-Swarm-camouflaged_gliding_salamander
franciscobdl/salamandra-estigiaV2
hector-gr/RLCR-v4-ks-batch-frontier-combo-cold-math
hector-gr/RLCR-v4-ks-highcov-volume-hotpot
hector-gr/RLCR-v4-ks-uniqueness-buf5k-noece-noaurc-cold-math
hector-gr/RLCR-v4-ks-uniqueness-cov0-entropy100-cold-math
inkw/llama3.1-8b-sft-sft-cmp-bt-merged
j05hr3d/Llama-3.2-1B-Instruct-C_M_T-SAM-AUX_CT_CE-RHO0_05
j05hr3d/Llama-3.2-3B-Instruct-C_M_T-AUX_CT_CE_CM-2EP
j05hr3d/Llama-3.2-3B-Instruct-C_M_T-AUX_CT_CE_CM-SAM
j05hr3d/Llama-3.2-3B-Instruct-C_M_T-Reh_Dolly
jayzou3773/Qwen1.5-MOE-sft-ESFT-intent
jingyeom/seal3.1.6n_7b
legionlm/sonnet-llama-3.2-3b
li-muyang/zephyr-7b-gemma-dpo
mehuldamani/sft-qwen-maze-v2
nekomajin/Qwen2.5-0.5B-Instruct-Gensyn-Swarm-mighty_hoarse_camel
norallm/normistral-7b-scratch
omrisap/nemotron-7B-12K
omrisap/nemotron-7B-6K
open-unlearning/unlearn_tofu_Llama-3.2-1B-Instruct_forget10_NPO_lr1e-05_beta0.5_alpha1_epoch10
AITeamVN/GRPO-VI-Qwen2-7B-RAG
Azazelle/SlimMelodicMaid
BlossomsAI/Qwen2.5-Coder-7B-Instruct-Uncensored
CYFRAGOVPL/Llama-PLLuM-8B-chat
ChengsenWang/ChatTime-1-7B-Chat
CobraMamba/mamba-gpt-7b-v1
DavidAU/Llama3.3-8B-Instruct-Thinking-Claude-4.5-Opus-High-Reasoning
Goekdeniz-Guelmez/Josiefied-Qwen2.5-7B-Instruct-abliterated-v2
GreatCaptainNemo/ProLLaMA
Q-bert/Optimus-7B
Yukang/LongAlpaca-7B
Zhengyi/LLaMA-Mesh
aari1995/germeo-7b-laser
abeja/ABEJA-Qwen2.5-7b-Japanese-v0.1
adamo1139/Yi-6B-200K-AEZAKMI-v2
adamo1139/aya-expanse-8b-ungated
adityakum667388/lumichats-v1.1
aiplanet/buddhi-128k-chat-7b
ajibawa-2023/Uncensored-Frank-7B
ajibawa-2023/Uncensored-Jordan-7B
alperiox/trendyol-7b-base-v1-mtLoRA_entr
alpha-ai/clinical-info-corrector-1B
Neko-Institute-of-Science/pygmalion-7b
Norquinal/llama-2-7b-claude-chat
Norquinal/llama-2-7b-claude-chat-rp
Novaciano/SEX_ROLEPLAY-3.2-1B
Open-Orca/Mistral-7B-OpenOrca
Open-Orca/Mistral-7B-SlimOrca
Orion-zhen/Qwen2.5-7B-Instruct-Uncensored
PhysicsWallahAI/Aryabhata-1.0
PleIAs/Baguettotron
Qwen/Qwen2.5-Coder-7B
Ramikan-BR/tinyllama-coder-py-v11
ReDiX/Qwen2.5-0.5B-Instruct-ITA
RoadQAQ/Qwen2.5-Math-1.5B-16k-think
Roc-M/M-project
Saxo/Linkbricks-Horizon-AI-Korean-Gemma-2-sft-dpo-27B
Weyaxi/Dolphin-Nebula-7B
Weyaxi/Platypus-Nebula-v2-7B
Weyaxi/SynthIA-v1.3-Nebula-v2-7B
Weyaxi/zephyr-alpha-Nebula-v2-7B
Weyaxi/zephyr-beta-Nebula-v2-7B
ringover/ringo-callanalysis-llama3.2-3b
the-jb/tofu_Llama-3.2-3B-Instruct_forget10_DPO
azalahmadkhan/Llama-3.2-3B-Instruct-GRPO-vanilla-G-8-novllm-25pct
the-jb/tofu_Llama-3.2-3B-Instruct_forget10_GradDiff
the-jb/tofu_Llama-3.2-3B-Instruct_forget10_NPO
the-jb/tofu_Llama-3.2-3B-Instruct_forget10_RMU
uyenlk/GA_forget10_1e-5_Llama-3.2-3B-Instruct
uyenlk/SimNPO_forget10_5e-5_Llama-3.2-3B-Instruct_gamma0.25_delta1.0_beta3.5
criscarleo/qwen2.5-coder-3b-abliterated
uyenlk/UNDIAL_forget10_5e-5_Llama-3.2-3B-Instruct_alpha1_beta30
GraphWiz/LLaMA2-7B
fearlessdots/WizardLM-2-7B-abliterated
kayrab/doktor-llama-3-cosmos-8b
luohy/SAIL-7b
vikash06/llama-2-7b-small-model-new
Nekochu/Confluence-Renegade-7B
SciPhi/SciPhi-Self-RAG-Mistral-7B-32k
Weyaxi/OpenOrca-Zephyr-7B
AtAndDev/CapybaraMarcoroni-7B
Azazelle/Silicon-Medley
Azazelle/smol_bruin-7b
BioMistral/BioMistral-7B-DARE
Cartinoe5930/Llama2_init_Mistral
Kamka-IT/shadow-clown-BioMistral-7B-DARE
RatanRohith/NeuralPizza-7B-V0.1
Sheikhaei/llama-3.2-1b-english-persian-translator
AliMaatouk/LLama-3-8B-Tele-it
Alienpenguin10/MAIN-M3PO-baseline-trial1-seed42
Stopwolf/Cerberus-7B-slerp
hyper-accel/tiny-random-llama
jayesh1234343/qwen-insurance-full
Athkal/model-sft-dare-resta
vsple/LegalBuddy-Pro-Final
Alienpenguin10/MAIN-M3PO-bahdanau-trial1-seed42
Aryanne/ereb-test
how3751/Planner_3B_1.0
xw1234gan/Extended_GRPO_KL_Qwen2.5-3B-Instruct_MATH_beta0_lr1e-05_mb2_ga128_n2048_seed42
xw1234gan/Extended_Merging_Prob_Qwen2.5-3B-Instruct_MATH_lr1e-05_mb2_ga128_n2048_seed42
xw1234gan/Main_fixed_MATH_3B_step_6
j05hr3d/Llama-3.2-3B-Instruct-C_M_T-AUX_INVERT
j05hr3d/Llama-3.2-3B-Instruct-C_M_T-SAM_RHO0_02-SEED1001
j05hr3d/Llama-3.2-3B-Instruct-C_M_T-SEED999
j05hr3d/Llama-3.2-3B-Instruct-C_M_T-AUX_INVERT-SEED999
j05hr3d/Llama-3.2-3B-Instruct-C_M_T_CT_CE_CM-2EP-SEED1001
lamm-mit/Graph-Preflexor_01062025
kingabzpro/llama-3.2-3b-it-Ecommerce-ChatBot
team-lucid/mptk-1b
j05hr3d/Llama-3.2-3B-Instruct-C_M_T-2EP
j05hr3d/Llama-3.2-3B-Instruct-C_M_T-DOLLY
AshtonIsNotHere/CodeLlama_7B_nlp_pp
DanielClough/Candle_Snorkel-Mistral-PairRM-DPO
jsfs11/MoEv4Config-TestWeightedTIES-7b
prince-canuma/Llama-3-6B-v0.1
d1shs0ap/cognitive-behaviors-Llama-3.2-3B
3uer/Qwen2.5-7B-Instruct-abliterated-v3
uukuguy/speechless-zephyr-code-functionary-7b
Undi95/Borealis-10.7B-DPO
websystemspl/Bielik-11B-v3.0-Instruct-128k
elyza/ELYZA-japanese-Llama-2-13b-fast
webbigdata/ALMA-7B-Ja-V2
turkerberkdonmez/TUSGPT-TR-Medical-9B
Bingsu/exaone-3.0-7.8b-it
ThijsL202/Merged_Roleplay_Dominant_Model_TEST
enstazao/Qalb-1.0-8B-Instruct
mesolitica/malaysian-llama-3-8b-instruct-16k
samuelsimko/Meta-Llama-3-8B-Instruct-Triplet
Inabia-AI/botox-ad-hoc-claims
Vaxispraxis/Llama-3.1-8B-Instruct-heretic
informatiker/Llama-3-8B-Instruct-abliterated
trendmicro-ailab/Llama-Primus-Merged
chentong00/Llama-3.1-8B-ParaPO
dmis-lab/llama-3.1-medprm-reward-v1.0
winglian/Llama-3-8b-64k-PoSE
hudson5/llama-3-8b-chat-doctor
Unbabel/Tower-Plus-9B
alnrg2arg/test
winglian/mistral-11b-128k
voidful/Llama-Breeze2-8B-Instruct-text-only
beomi/OPEN-SOLAR-KO-10.7B
sanghwa-na/llama2-13b.kor
chargoddard/llama-polyglot-13b
royallab/PsyOrca2-13b-DARE
BlueNipples/TimeCrystal-l2-13B
guardrail/llama-2-7b-guanaco-instruct-sharded
hamxea/Llama-2-7b-chat-hf-activity-fine-tuned-v3
daniellnichols/spack-llama-7b
somukandula/cx-filler-model
iamshnoo/combined_no_africa_with_metadata_1b_step2k
PrasannaPaithankar/qwen2.5-1.5b-sft-dare-resta
allout2726/model_sft_dare
holi-lab/qwen-2.5-1.5b-multiwoz-finetuned_fp16
thrnn/qwen2.5-1.5b-sft-dare-resta
NoseGlasses/qwen25_1_5b_korean_unsloth
Soohyunai/qwen25_1_5b_korean_unsloth
popman12/qwen25_1_5b_korean_unsloth
quangne/text2diagram-AceMath-1.5B-Instruct-merged-1k
thrnn/qwen2.5-1.5b-medical-sft-dare
eekay/gemma-2b-it-steer-dog-numbers-ft
shibing624/chinese-alpaca-plus-7b-hf
cjvt/GaMS-27B-Instruct-Nemotron
sweepai/sweep-next-edit-v2-7B
yufeng1/OpenThinker-7B-type6-e5-max-b64-alpha0_28125-2
xw1234gan/Merging_Prob_Qwen2.5-7B-Instruct_MATH_lr1e-05_mb2_ga128_n2048_seed42
xw1234gan/Main_fixed_MATH_1_5B_BaseAnchor_step_2
xw1234gan/Main_fixed_MATH_1_5B_BaseAnchor_step_1
sstoica12/acquisition_llama-3_1-8b_bins_numina_confidence
sathiiiii/polyalign-qwen2.5-3b-en-dist-sft
open-sci/sft__ot30k_SmolLM2-1.7B-Instruct-16k
open-sci/sft__ot30k_SmolLM2-1.7B-16k-SFT-Tulu3-decontaminated
open-sci/sft__ot30k_Qwen2.5-1.5B-DPO-Tulu3-decontaminated
nilarnabdebnath/Qwen2.5-1.5B-Instruct_gsm8k
mehuldamani/bug_fixing_new-arl-add_multiply
ishikaa/acquisition_student_qwen3bins_numina_proximity
ishikaa/acquisition_student_qwen3bins_numina_diversity
ishikaa/acquisition_student_qwen3bins_numina_confidence
ishikaa/acquisition_qwen3bins_numina_proximity
ishikaa/acquisition_qwen3bins_numina_gradient
ishikaa/acquisition_qwen3bins_numina_format
ishikaa/acquisition_qwen3bins_numina_diversity
ishikaa/acquisition_qwen3bins_numina_confidence
ishikaa/acquisition_qwen3bins_medmcqa_format
W-61/llama-3-8b-base-new-dpo-hh-harmless-s_star0.6-4xh200-batch-64-20260421-213851
Te-REx/Qwen-1.5B-Customer-Support
Tapask/telecom-oss-8b-merged
Skronak/min0-translator-v1
Ramikan-BR/TiamaPY-v34
RafikContractzlab/qwen2.5-3b-sft
NotoriousH2/A.X-4.0-Light-Sunbi-Merged
Neelectric/Llama-3.1-8B-Instruct_SafeGrad_mathv00.06
KickItLikeShika/Qwen2.5-1.5B-Instruct-SFT-GRPO-GSM8K
JdoubleU/careconnect-llama3.2-3b
xw1234gan/SMOKE_Merging_Prob_Qwen2.5-7B-Instruct_MATH_lr1e-05_mb2_ga4_n16_seed42
xw1234gan/Main_fixed_MATH_7B_step_9
xw1234gan/Main_fixed_MATH_7B_step_8
xw1234gan/Main_fixed_MATH_7B_step_7
xw1234gan/Main_fixed_MATH_7B_step_6
xw1234gan/Main_fixed_MATH_7B_step_4
xw1234gan/Main_fixed_MATH_7B_step_2
xw1234gan/Main_fixed_MATH_7B_step_10
xw1234gan/Main_fixed_MATH_7B_step_1
tengfeima-ai/Qwen2.5-0.5B-Math-SFT-1024
tengfeima-ai/Qwen2.5-0.5B-Math-GRPO-Concise
seopbo/zerorlvrmath-qwen2.5-1.5b
seopbo/sft-qwen2.5-1.5b
seopbo/rlvrmath-qwen2.5-1.5b
seopbo/rlvrcode-qwen2.5-1.5b
paudelnirajan/general-kd-Qwen2.5-0.5B-Instruct-ber-5000-4500
paudelnirajan/general-kd-Qwen2.5-0.5B-Instruct-ber-5000-4000
paudelnirajan/general-kd-Qwen2.5-0.5B-Instruct-ber-5000-3500
olabhinavlo/demosample
muverqqw/Noir-Lightning
muverqqw/Noir
malFlexion/the-legacy-lora-merged
laihuiyuan/mCoT
hyper-accel/ci-random-bfloat16-llama2-7b
how3751/Coder_7B_1.0