Rishichaitanya-Nalluri/zengrad-tensorflow

No problems traced to this repository code in these checks.

Reviewed as expected for a test run: attempted writes that failed, so nothing was written.

OSV scanned supported lockfiles; unpinned dependencies and unsupported formats are outside this check.

A short run cannot establish that code is safe.

Connections show IP addresses; proxies can hide the final hostname.

Tracing covers dispatched commands and their children, not the whole OS; guest root can interfere with tracing.

Full safety evidence →

• The README Implementation section imports ZenGrad from zengrad_tensorflow and it performs a Keras optimizer update with the documented hyperparameters: Real input produced the expected result: [0.9981059513768245, 2.003743288932412]

  Source: from zengrad_tensorflow import ZenGrad # Import the ZenGrad optimizer

  Input: learning_rate=0.01, weight_decay=0.0001, momentum=0.9, epsilon=1e-8; weights=[1.0,2.0], gradient=[0.1,-0.2]

  Expected: updated weights [0.9981059513768245,2.003743288932412]

  Command: Python API zengrad_tensorflow.ZenGrad

  Observed: :InitializeLog() is called are written to STDERR I0000 00:00:1793173794.776307 3047 port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`. I0000 00:00:1793173794.788852 3047 cudart_stub.cc:31] Could not find cuda drivers on your machine, GPU will not be used. E0000 00:00:1793173796.187311 3047 cuda_platform.cc:52] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303)

Local or hosted: local

Run the ZenGrad TensorFlow optimizer (zengrad_tensorflow.ZenGrad) on CPU for one Keras apply_gradients update with the README hyperparameters, and check the updated weights against the expected values.

cd <guest-path> && pwd && ls -la && cat README.md | head -c 7000 && python3 --version
cd <guest-path> && ls -la zengrad_tensorflow .coldpass && cat setup.py && wc -c zengrad_tensorflow/* && cat zengrad_tensorflow/*.py | head -c 6000; ls /inputs 2>/dev/null; cat <guest-path> 2>/dev/null | head -5
mkdir -p <guest-path> && python3 -m venv <guest-path> && <guest-path> install -q --upgrade pip 2>&1 | tail -3; <guest-path> --version
cd /tmp && timeout 3000 <guest-path> install zengrad-tensorflow > <guest-path> 2>&1; echo rc=$?; tail -5 <guest-path>
cd <guest-path> && git init -q && git remote add origin https://github.com/Rishichaitanya-Nalluri/zengrad-tensorflow && timeout 300 git fetch --depth 1 origin cdd19353b333c973c65f8e25499c701ba3de0680 2>&1 | tail -3; echo rc=$?; git reset --mixed FETCH_HEAD 2>&1 | tail -2; git log --oneline -1; git status --short | head; <guest-path> -c "import zengrad_tensorflow,inspect;print(zengrad_tensorflow.__file__)"
cd <guest-path> && ls -la checker /inputs 2>&1 | head; grep -n "python_api\|tensorflow_step\|bytes_hex\|expected_value" checker/check.py | head -40
cd <guest-path> && sed -n 905,960p checker/check.py
cd <guest-path> && grep -n "PYTHON_EXAMPLE_RUNNER\|sys.executable\|def run_python\|VIRTUAL_ENV\|venv" checker/check.py | head -30
which python3; python3 -m pip --version 2>&1 | head -2; python3 -c "import tensorflow" 2>&1 | tail -1
cd /tmp && timeout 3000 python3 -m pip install --user --break-system-packages -q zengrad-tensorflow > <guest-path> 2>&1; echo rc=$?; tail -3 <guest-path>; cd /tmp && python3 -c "import tensorflow as tf, keras; print(tf.__version__, keras.__version__)" 2>&1 | tail -1
cd <guest-path> && cat > probe.json <<'EOF'
{"kind":"python_api","module":"zengrad_tensorflow","call":"ZenGrad","args":[],"kwargs":{"learning_rate":0.01,"weight_decay":0.0001,"momentum":0.9,"epsilon":1e-8},"tensorflow_step":{"weights":[1.0,2.0],"gradient":[0.1,-0.2]},"expected_value":[0.9981059513768245,2.003743288932412],"cwd":"<guest-path>"}
EOF
cat > claims.json <<'EOF'
[{"kind":"claim","claim":"The README Implementation section imports ZenGrad from zengrad_tensorflow and it performs a Keras optimizer update with the documented hyperparameters","source":"from zengrad_tensorflow import ZenGrad  # Import the ZenGrad optimizer","input":"learning_rate=0.01, weight_decay=0.0001, momentum=0.9, epsilon=1e-8; weights=[1.0,2.0], gradient=[0.1,-0.2]","expected":"updated weights [0.9981059513768245,2.003743288932412]","probe_kind":"python_api","module":"zengrad_tensorflow","call":"ZenGrad","args":[],"kwargs":{"learning_rate":0.01,"weight_decay":0.0001,"momentum":0.9,"epsilon":1e-8},"tensorflow_step":{"weights":[1.0,2.0],"gradient":[0.1,-0.2]},"expected_value":[0.9981059513768245,2.003743288932412],"cwd":"<guest-path>"}]
EOF
cat > cycles.json <<'EOF'
{"detected":false,"checks":[],"reason":"Inspected README.md and the zengrad_tensorflow package (__init__.py, zengrad_tensorflow.py) and setup.py. ZenGrad is an optimizer class with no cron, loop, webhook, scheduler, --once or --now trigger, so there is no product cycle to run."}
EOF
cat > examples.json <<'EOF'
[]
EOF
for f in probe claims cycles examples; do <guest-path> -m json.tool $f.json >/dev/null && echo $f ok; done
cd /tmp && <env-var>=<guest-path> python3 -I <guest-path> --validate --replay 2>&1 | tail -c 4000
cd /tmp && <env-var>=<guest-path> python3 -I <guest-path> --validate --replay > <guest-path> 2>&1; echo rc=$?; python3 -c "
import json,sys
t=open('<guest-path>').read()
i=t.find('{')
d=json.loads(t[i:])
print({k:(v if not isinstance(v,(dict,list)) else '...') for k,v in d.items()})
" 2>&1 | head -c 2000
cd /tmp && python3 -c "
import json
t=open('<guest-path>').read()
d=json.loads(t[t.find('{'):])
print('problems:',d['problems'])
r=d['replay']; print('replay keys:', list(r.keys()) if isinstance(r,dict) else r)
" | head -c 3000; ls -la <guest-path>
cd /tmp && python3 -c "
import json
t=open('<guest-path>').read()
d=json.loads(t[t.find('{'):])
print(d['replay']['status'], d['replay']['details'] if not isinstance(d['replay']['details'],str) else d['replay']['details'][:300])
" | head -c 1500
cd <guest-path> && cat > .coldpass/report.json <<'EOF'
{"outcome":"working","fault_owner":null,"operation":"Run the ZenGrad TensorFlow optimizer (zengrad_tensorflow.ZenGrad) on CPU for one Keras apply_gradients update with the README hyperparameters, and check the updated weights against the expected values.","operation_scope":"CPU only, no GPU or account. Pinned source commit cdd19353b333c973c65f8e25499c701ba3de0680 (git init, remote origin, fetch --depth 1, reset --mixed). Python 3.12.3, tensorflow 2.21.0, keras 3.15.1. Prerequisites: pip install zengrad-tensorflow (PyPI 0.1.1) in a venv at <guest-path> and in system python3 (user install) because the checker runs the python_api through system python3.","steps":["git init; git remote add origin https://github.com/Rishichaitanya-Nalluri/zengrad-tensorflow; git fetch --depth 1 origin cdd19353b333c973c65f8e25499c701ba3de0680; git reset --mixed FETCH_HEAD","python3 -m venv <guest-path>; <guest-path> install zengrad-tensorflow (README Installation)","python3 -m pip install --user --break-system-packages zengrad-tensorflow (system interpreter for the checker)","README Implementation: from zengrad_tensorflow import ZenGrad; ZenGrad(learning_rate=0.01, weight_decay=1e-4, momentum=0.9, epsilon=1e-8); python_api replay with tensorflow_step weights=[1.0,2.0], gradient=[0.1,-0.2]"],"deviations":["The README model.compile snippet is a placeholder (model = ... and .......), so it was not run as-is; the optimizer was exercised directly through apply_gradients with a tf.Variable, as the checker's tensorflow_step does.","Installed the package from PyPI per the README rather than from the pinned commit; the import during the probe resolves to the pinned workspace source, which the checker verifies.","Actual output [0.9981059514686015, 2.003743288752119] differs from the expected [0.9981059513768245, 2.003743288932412] by about 1e-10 to 2e-10; the checker's tolerance accepted it.","CUDA cuInit warnings in TensorFlow output are expected on a machine without GPU and do not affect the CPU run.","No runnable documented example or test was found; examples.json is empty. No cycle: detected=false."],"stuck_step":null,"main_blocker":null,"plain_failure":null,"suggested_fix":"None required. Optionally fix the README snippet (CRLF, placeholder model) to a runnable example."}
EOF
<guest-path> -m json.tool .coldpass/report.json >/dev/null && echo ok; cd /tmp && <env-var>=<guest-path> python3 -I <guest-path> --validate --replay 2>&1 | tail -c 300