wuyang
90 lines
2.2 KiB
JSON
90 lines
2.2 KiB
JSON
{
|
|
"schema_version": 1,
|
|
"protocol_id": "llm-atlas-k3-attnres-forward-training-v1",
|
|
"status": "frozen-before-model-output",
|
|
"parent_protocol_id": "llm-atlas-k3-attnres-gradient-scale-v1",
|
|
"parent_manifest": "experiments/k3/attnres_gradient/manifest.json",
|
|
"architecture": "block",
|
|
"depth": 32,
|
|
"formal_steps": 8000,
|
|
"smoke_steps": 20,
|
|
"batch_size": 32,
|
|
"formal_seeds": [
|
|
2026073001,
|
|
2026073002,
|
|
2026073003
|
|
],
|
|
"diagnostic_steps": [
|
|
0,
|
|
100,
|
|
500,
|
|
2000,
|
|
4000,
|
|
8000
|
|
],
|
|
"primary_step": 8000,
|
|
"spike_layers_1based": [21, 22, 23, 24, 25],
|
|
"historical_reference": {
|
|
"directory": "experiments/k3/attnres_gradient/results/raw",
|
|
"filename_template": "formal-depth-32-block-seed-{seed}.json",
|
|
"identity": "historical-paired-reference-not-contemporaneous-randomized-control"
|
|
},
|
|
"variants": {
|
|
"uniform_group_6_forward": {
|
|
"selected_depth_indices": [40, 41, 42, 43, 44, 45, 46, 47]
|
|
},
|
|
"uniform_group_7_forward": {
|
|
"selected_depth_indices": [48, 49, 50, 51, 52, 53, 54, 55]
|
|
},
|
|
"uniform_groups_6_7_forward": {
|
|
"selected_depth_indices": [40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55]
|
|
},
|
|
"uniform_group_7_mlp_forward": {
|
|
"selected_depth_indices": [49, 51, 53, 55]
|
|
}
|
|
},
|
|
"smoke_only_variants": [
|
|
"learned_reference"
|
|
],
|
|
"primary_variant": "uniform_groups_6_7_forward",
|
|
"replay": {
|
|
"variant": "uniform_groups_6_7_forward",
|
|
"seed": 2026073001
|
|
},
|
|
"selected_source_counts": {
|
|
"40": 6,
|
|
"41": 7,
|
|
"42": 7,
|
|
"43": 7,
|
|
"44": 7,
|
|
"45": 7,
|
|
"46": 7,
|
|
"47": 7,
|
|
"48": 7,
|
|
"49": 8,
|
|
"50": 8,
|
|
"51": 8,
|
|
"52": 8,
|
|
"53": 8,
|
|
"54": 8,
|
|
"55": 8
|
|
},
|
|
"thresholds": {
|
|
"positive_denominator_epsilon": 1e-30,
|
|
"material_relative_drop": 0.2,
|
|
"final_bpc_delta_per_seed_maximum": 0.05,
|
|
"final_bpc_delta_mean_maximum": 0.03,
|
|
"uniform_weight_max_abs_error": 1e-12,
|
|
"loss_scale_ratio_abs_error": 1e-5,
|
|
"loss_scale_shape_abs_error": 1e-6
|
|
},
|
|
"new_target_bytes": {
|
|
"per_cell": 65536000,
|
|
"formal_12_cells": 786432000,
|
|
"primary_replay": 65536000,
|
|
"total": 851968000
|
|
},
|
|
"historical_reference_target_bytes": 196608000,
|
|
"concurrency_maximum": 2
|
|
}
|