Spaces:
Running
Running
File size: 6,296 Bytes
9d98328 | 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 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 | """Draft canonical learning/synapse contract descriptor.
This module inventories current CPU semantics and known CUDA-reference
differences. Presence of the descriptor does not authorize cross-backend
learning equivalence or scientific use.
"""
from __future__ import annotations
import json
from dataclasses import dataclass
from hashlib import sha256
from typing import Final
LEARNING_CONTRACT_ID: Final[str] = "mhrn-learning-synapse-v0.1-draft"
LEARNING_CONTRACT_STATUS: Final[str] = "DRAFT_NOT_FROZEN"
LEARNING_CPU_REFERENCE: Final[str] = "src.learning.learning_engine.LearningEngine"
LEARNING_CUDA_STATUS: Final[str] = "NON_CANONICAL_DRAFT"
def learning_contract_descriptor() -> dict[str, object]:
"""Return the current normative inventory without claiming a freeze."""
return {
"classification": "LEARNING_SYNAPSE_CONTRACT_DESCRIPTOR",
"scientific_evidence": False,
"contract_id": LEARNING_CONTRACT_ID,
"contract_status": LEARNING_CONTRACT_STATUS,
"cpu_reference": LEARNING_CPU_REFERENCE,
"cuda_status": LEARNING_CUDA_STATUS,
"event_order": "SORTED_(PRE_ID,TARGET_ID)_AFTER_COMPLETED_NETWORK_TICK",
"same_tick_pair_rule": "ZERO_PAIR_CONTRIBUTION",
"stdp": {
"rule": "NEAREST_NEIGHBOUR_PAIR",
"ltp": "+a_plus*exp(-dt/tau_plus), dt>0",
"ltd": "-a_minus*exp(dt/tau_minus), dt<0",
"direct_weight_clamp": True,
},
"eligibility": {
"rule": "LAZY_EXPONENTIAL_DECAY_THEN_ADD_PAIR_DELTA",
"decay": "exp(-dt/tau_ticks)",
"checkpointed_cpu_state": True,
},
"reward": {
"effective_tick": "emitted_tick+reward_delay_ticks",
"rule": "learning_rate*reward*eligibility(effective_tick)",
"due_order": "PENDING_REWARD_INSERTION_ORDER",
"synapse_order": "SORTED_STABLE_KEY",
"optional_trace_reset": True,
},
"delayed_signal": {
"required_rule": "EMITTED_DELAYED_AMPLITUDE_IS_IMMUTABLE",
"cross_backend_alignment": False,
},
"stp_candidate": {
"status": "CUDA_REFERENCE_CANDIDATE_NOT_FROZEN",
"release_rng": "COUNTER_RNG(seed,tick,edge)",
"release_probability": "min(0.95,0.25+0.7*available)",
"depletion": "max(0.1,available*0.72)",
"recovery": "min(1.0,available+0.025)",
},
"known_semantic_gaps": [
"CPU_STABLE_IDENTITY_IS_(PRE_ID,TARGET_ID)_AND_DOES_NOT_CANONICALIZE_PARALLEL_EDGES",
"CPU_REFERENCE_HAS_NO_CANONICAL_STP_STATE",
"CUDA_REFERENCE_USES_FIXED_CREDIT_WINDOW_WHILE_CPU_REFERENCE_USES_DELAYED_REWARD_PLUS_ELIGIBILITY",
"CUDA_REFERENCE_APPLIES_PER_TICK_WEIGHT_DECAY_WHILE_CPU_REFERENCE_DOES_NOT",
"CUDA_REFERENCE_PAIR_AMPLITUDES_DIFFER_FROM_CPU_DEFAULTS",
"CPU_CUDA_LEARNING_SEMANTICS_NOT_ALIGNED",
],
}
def learning_contract_hash() -> str:
payload = json.dumps(
learning_contract_descriptor(),
ensure_ascii=True,
sort_keys=True,
separators=(",", ":"),
).encode("utf-8")
return sha256(payload).hexdigest()
@dataclass(frozen=True, slots=True)
class LearningContractState:
contract_frozen: bool = False
stable_edge_identity_complete: bool = False
stp_semantics_frozen: bool = False
reward_credit_semantics_aligned: bool = False
weight_decay_semantics_aligned: bool = False
cuda_semantics_aligned: bool = False
checkpoint_roundtrip_available: bool = True
@property
def ready_for_cross_backend_learning(self) -> bool:
return (
self.contract_frozen
and self.stable_edge_identity_complete
and self.stp_semantics_frozen
and self.reward_credit_semantics_aligned
and self.weight_decay_semantics_aligned
and self.cuda_semantics_aligned
and self.checkpoint_roundtrip_available
)
def to_mapping(self) -> dict[str, object]:
blockers: list[str] = []
if not self.contract_frozen:
blockers.append("LEARNING_CONTRACT_NOT_FROZEN")
if not self.stable_edge_identity_complete:
blockers.append("STABLE_EDGE_ID_NOT_CANONICAL")
if not self.stp_semantics_frozen:
blockers.append("STP_SEMANTICS_NOT_FROZEN")
if not self.reward_credit_semantics_aligned:
blockers.append("REWARD_CREDIT_SEMANTICS_NOT_ALIGNED")
if not self.weight_decay_semantics_aligned:
blockers.append("WEIGHT_DECAY_SEMANTICS_NOT_ALIGNED")
if not self.cuda_semantics_aligned:
blockers.append("CUDA_LEARNING_SEMANTICS_NOT_ALIGNED")
if not self.checkpoint_roundtrip_available:
blockers.append("LEARNING_CHECKPOINT_ROUNDTRIP_NOT_AVAILABLE")
return {
"classification": "LEARNING_SYNAPSE_CONTRACT_STATUS",
"scientific_evidence": False,
"contract_id": LEARNING_CONTRACT_ID,
"contract_status": LEARNING_CONTRACT_STATUS,
"descriptor_hash": learning_contract_hash(),
"cpu_reference": LEARNING_CPU_REFERENCE,
"cuda_status": LEARNING_CUDA_STATUS,
"ready_for_cross_backend_learning": self.ready_for_cross_backend_learning,
"blockers": blockers,
}
def learning_contract_status() -> LearningContractState:
return LearningContractState()
def learning_contract_check() -> bool:
first = learning_contract_hash()
second = learning_contract_hash()
state = learning_contract_status()
mapping = state.to_mapping()
blockers = mapping["blockers"]
return (
isinstance(blockers, list)
and first == second
and len(first) == 64
and not state.ready_for_cross_backend_learning
and mapping["scientific_evidence"] is False
and "LEARNING_CONTRACT_NOT_FROZEN" in blockers
)
__all__ = [
"LEARNING_CONTRACT_ID",
"LEARNING_CONTRACT_STATUS",
"LEARNING_CPU_REFERENCE",
"LEARNING_CUDA_STATUS",
"LearningContractState",
"learning_contract_check",
"learning_contract_descriptor",
"learning_contract_hash",
"learning_contract_status",
]
|