death-mountain-sp-1b

Death Mountain PPO policy: sp-1b

Download from Hugging Face

python -m pip install huggingface_hub
hf download rvorias/death-mountain-sp-1b --revision v1.0 --local-dir death-mountain-sp-1b
cd death-mountain-sp-1b
python -m pip install -r requirements.txt
python policy_api.py --self-test

This model uses the included DeathMountainPolicy API. It is a custom PyTorch policy, rather than a Transformers AutoModel checkpoint. The included weights and benchmark records match the verified GitHub release.

The retained project champion is now available as MIT-licensed weights and standalone CPU inference code. On 16,000 fresh games it scored 359.80 mean XP and 207 median XP, with zero truncations.

A subsequent 199.75M-transition, four-run training comparison found no candidate that passed the predeclared promotion rule. This release therefore preserves the incumbent. Model card and results.

python -m pip install -r requirements.txt
python policy_api.py --self-test

The self-test uses the included real-observation fixture and expected outputs; it needs no game installation. To play, supply observations and legal masks from a compatible environment. Observation/action contract.

Inference API

No game engine, C compiler, dmfast, Triton, CUDA driver, or original source checkout is needed. Install requirements.txt, then provide observations and legal-action masks from your environment. When exported with --validate, run the included real-observation self-test without a simulator:

python policy_api.py --self-test

The test checks artifact hashes, exact-feature bitwise equality, and raw-input equality within atol=1e-4/rtol=1e-5 against saved CPU outputs. Both tests cover logits, value, and recurrent state. Bitwise reproduction targets the recorded package versions and CPU numerical behavior; different PyTorch versions or CPU backends may require investigating numerical differences. The numeric interface is 463 raw float32 observation fields and 57 bool action-mask fields. Observations must include the existing combat-simulation fields; this model does not compute game dynamics.

import torch
from policy_api import DeathMountainPolicy

policy = DeathMountainPolicy()
state = policy.initial_state(batch_size)
# raw_obs: (batch_size, 463); legal_mask: (batch_size, 57)
logits, value, state = policy.step(raw_obs, legal_mask, state)
actions = torch.distributions.Categorical(logits=logits).sample()
# After your environment advances, reset memory for ended episodes.
state = policy.reset_state(state, done)

The default excludes macro actions 7–10 with the canonical nonempty-mask fallback. Pass primitive_actions=False to use exactly your supplied legal mask. Use logits.argmax(-1) for greedy actions; PPO score reports ordinarily use sampling. A NumPy reference derives semantic features from raw observations. Its floating-point rounding can differ slightly from native C features. An optional semantic_features argument accepts the exact 88/90 features from an external evaluator for exact checkpoint equivalence. The raw-input validation report records actual numerical differences; it does not claim bit-identical sampled trajectories or certify benchmark scores across implementations.

Weights are safetensors without optimizer state; full architecture and original checkpoint metadata are retained. See manifest.json for hashes and LICENSE for MIT terms.