Instructions to use bowmanslayer/VeriLoop-E2-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use bowmanslayer/VeriLoop-E2-GGUF with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf bowmanslayer/VeriLoop-E2-GGUF:F16 # Run inference directly in the terminal: llama cli -hf bowmanslayer/VeriLoop-E2-GGUF:F16
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf bowmanslayer/VeriLoop-E2-GGUF:F16 # Run inference directly in the terminal: llama cli -hf bowmanslayer/VeriLoop-E2-GGUF:F16
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-app # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama serve -hf bowmanslayer/VeriLoop-E2-GGUF:F16 # Run inference directly in the terminal: ./build/bin/llama cli -hf bowmanslayer/VeriLoop-E2-GGUF:F16
Use Docker
docker model run hf.co/bowmanslayer/VeriLoop-E2-GGUF:F16
- Llama new
- LM Studio
- Jan
- vLLM
How to use bowmanslayer/VeriLoop-E2-GGUF with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "bowmanslayer/VeriLoop-E2-GGUF" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "bowmanslayer/VeriLoop-E2-GGUF", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/bowmanslayer/VeriLoop-E2-GGUF:F16
- Ollama
How to use bowmanslayer/VeriLoop-E2-GGUF with Ollama:
ollama run hf.co/bowmanslayer/VeriLoop-E2-GGUF:F16
- Unsloth Desktop
- Pi
How to use bowmanslayer/VeriLoop-E2-GGUF with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf bowmanslayer/VeriLoop-E2-GGUF:F16
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:9931/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "bowmanslayer/VeriLoop-E2-GGUF:F16" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use bowmanslayer/VeriLoop-E2-GGUF with Docker Model Runner:
docker model run hf.co/bowmanslayer/VeriLoop-E2-GGUF:F16
- Lemonade
How to use bowmanslayer/VeriLoop-E2-GGUF with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull bowmanslayer/VeriLoop-E2-GGUF:F16
Run and chat with the model
lemonade run user.VeriLoop-E2-GGUF-F16
List all available models
lemonade list
- Hermes Agent
How to use bowmanslayer/VeriLoop-E2-GGUF with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf bowmanslayer/VeriLoop-E2-GGUF:F16
Configure Hermes
# Install Hermes: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash hermes setup # Point Hermes at the local server: hermes config set model.provider custom hermes config set model.base_url http://127.0.0.1:9931/v1 hermes config set model.default bowmanslayer/VeriLoop-E2-GGUF:F16
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use bowmanslayer/VeriLoop-E2-GGUF with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf bowmanslayer/VeriLoop-E2-GGUF:F16
Configure OpenClaw
# Install OpenClaw: npm install -g openclaw@latest # Register the local server and set it as the default model: openclaw onboard --non-interactive --mode local \ --auth-choice custom-api-key \ --custom-base-url http://127.0.0.1:9931/v1 \ --custom-model-id "bowmanslayer/VeriLoop-E2-GGUF:F16" \ --custom-provider-id llama-cpp \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
File size: 4,189 Bytes
1816fff | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 | """Frozen legacy rubric with engine field fixes, frozen questions and isolated code execution."""
import pathlib,json,subprocess,uuid,os
import legacy_capability as h
from protocol import normalize_content
ROOT=pathlib.Path(__file__).parent
load_legacy=h.load_items
def load_items(task,n,seed):
data=json.loads((ROOT/'questions.json').read_text())
assert seed==data['seed']
items=data['tasks'][task]
assert len(items)==n,(task,len(items),n)
return items
h.load_items=load_items
# Expose vLLM's current reasoning field and finish reason without changing scoring.
old_call=h.call
def call(session,args,item,spec):
class Adapter:
def post(self,*a,**kw):
r=session.post(*a,**kw)
if r.ok:
d=r.json();c=d['choices'][0];m=c['message']
raw=m.get('content') or ''
content,reasoning,mode=normalize_content(raw,c.get('finish_reason')) if args.thinking else (raw,'','thinking-disabled')
m['content']=content
m['reasoning_content']=reasoning
d.setdefault('usage',{})['content_mode']=mode
d.setdefault('usage',{})['finish_reason']=c.get('finish_reason')
r.json=lambda:d
return r
return old_call(Adapter(),args,item,spec)
h.call=call
def score_code(text,gold,timeout=15):
code=h._extract_code(text)
if not code.strip():return False
program='import math,re,sys,itertools,collections,heapq,functools,string\n'+code+'\n'+gold['test']+'\ncheck('+gold['entry_point']+')\n'
name='veriloop-eval-'+uuid.uuid4().hex
cmd=['docker','run','--rm','-i','--name',name,'--network','none','--read-only','--cap-drop','ALL','--security-opt','no-new-privileges','--pids-limit','64','--memory','256m','--cpus','1','--user','65534:65534','--entrypoint','python3','flashnext-base:locked-20261008','-I','-']
try:return subprocess.run(cmd,input=program,text=True,stdout=subprocess.DEVNULL,stderr=subprocess.DEVNULL,timeout=timeout).returncode==0
except subprocess.TimeoutExpired:return False
finally:subprocess.run(['docker','rm','-f',name],stdout=subprocess.DEVNULL,stderr=subprocess.DEVNULL)
h.score_code=score_code
def main():
import argparse,threading,time
from concurrent.futures import ThreadPoolExecutor
p=argparse.ArgumentParser()
for flag in ['base-url','model','label','outdir']:p.add_argument('--'+flag,required=True)
p.add_argument('--tasks',default='all');p.add_argument('--seed',type=int,default=20260816)
p.add_argument('--max-tokens',dest='max_tokens_override',type=int,default=32768)
p.add_argument('--concurrency',type=int,default=16);p.add_argument('--timeout',type=int,default=3600)
p.add_argument('--no-thinking',dest='thinking',action='store_false')
args=p.parse_args();args.n_override=0;args.temperature=0.;args.top_p=1.;args.retries=3
tasks=h.ALL_TASKS if args.tasks=='all' else args.tasks.split(',')
limit=threading.BoundedSemaphore(args.concurrency)
underlying=h.call
def bounded(*a,**kw):
with limit:return underlying(*a,**kw)
h.call=bounded
out=pathlib.Path(args.outdir);out.mkdir(exist_ok=True)
summary={'label':args.label,'model':args.model,'base_url':args.base_url,'seed':args.seed,'thinking':args.thinking,'started':time.strftime('%F %T'),'tasks':{},'scheduler':'task workers share one request semaphore'}
def task(t):
ckpt=out/(args.label+'-'+t+'.json')
if ckpt.exists():r=json.loads(ckpt.read_text())
else:
print('START',t,flush=True);r=h.run_task(args,t)
if r['errors']:raise RuntimeError(t+' request errors')
ckpt.write_text(json.dumps(r,ensure_ascii=False,indent=2))
print('RESULT',t,r['hits'],r['n'],'empty',r.get('empty'),'length',r.get('length_limited'),flush=True)
return t,{k:v for k,v in r.items() if k!='details'}
with ThreadPoolExecutor(max_workers=len(tasks)) as pool:
for t,r in pool.map(task,tasks):summary['tasks'][t]=r
summary['finished']=time.strftime('%F %T')
(out/('SUMMARY-'+args.label+'.json')).write_text(json.dumps(summary,indent=2))
if __name__=='__main__':main()
|