petra345 commited on
Commit
96eb08d
·
verified ·
1 Parent(s): 5dd0ed9

Risk-reviewed release candidate payload

Browse files
decision_table.csv CHANGED
@@ -1,8 +1,8 @@
1
  rank,checkpoint,eval_accuracy,toxicity_rate,latency_ms,core_reasoning_avg,language_understanding_avg,generation_tasks_avg,specialized_capabilities_avg,robustness_floor,risk_adjusted_score,weakest_group,model_sha256_12
2
- 1,checkpoints/step_800,0.752,0.018,185.000,0.719,0.753,0.686,0.774,0.686,0.672,Generation Tasks,01934db10231
3
- 2,checkpoints/step_700,0.739,0.016,175.000,0.716,0.751,0.682,0.770,0.682,0.669,Generation Tasks,beeb0044e3ea
4
  3,checkpoints/step_600,0.716,0.019,205.000,0.712,0.750,0.678,0.766,0.678,0.663,Generation Tasks,177138f2d2c3
5
- 4,checkpoints/step_900,0.771,0.033,260.000,0.726,0.755,0.677,0.776,0.677,0.656,Generation Tasks,d837916e96de
6
  5,checkpoints/step_500,0.691,0.021,210.000,0.704,0.736,0.670,0.750,0.670,0.654,Generation Tasks,49e27938931c
7
  6,checkpoints/step_400,0.668,0.022,230.000,0.696,0.722,0.657,0.733,0.657,0.640,Generation Tasks,71780498800f
8
  7,checkpoints/step_300,0.641,0.024,240.000,0.684,0.710,0.644,0.721,0.644,0.626,Generation Tasks,f549bc42f96d
 
1
  rank,checkpoint,eval_accuracy,toxicity_rate,latency_ms,core_reasoning_avg,language_understanding_avg,generation_tasks_avg,specialized_capabilities_avg,robustness_floor,risk_adjusted_score,weakest_group,model_sha256_12
2
+ 1,checkpoints/step_800,0.752,0.018,185.000,0.719,0.754,0.686,0.774,0.686,0.672,Generation Tasks,01934db10231
3
+ 2,checkpoints/step_700,0.739,0.016,175.000,0.716,0.751,0.682,0.771,0.682,0.669,Generation Tasks,beeb0044e3ea
4
  3,checkpoints/step_600,0.716,0.019,205.000,0.712,0.750,0.678,0.766,0.678,0.663,Generation Tasks,177138f2d2c3
5
+ 4,checkpoints/step_900,0.771,0.033,260.000,0.726,0.756,0.677,0.777,0.677,0.656,Generation Tasks,d837916e96de
6
  5,checkpoints/step_500,0.691,0.021,210.000,0.704,0.736,0.670,0.750,0.670,0.654,Generation Tasks,49e27938931c
7
  6,checkpoints/step_400,0.668,0.022,230.000,0.696,0.722,0.657,0.733,0.657,0.640,Generation Tasks,71780498800f
8
  7,checkpoints/step_300,0.641,0.024,240.000,0.684,0.710,0.644,0.721,0.644,0.626,Generation Tasks,f549bc42f96d
group_summary.csv CHANGED
@@ -1,5 +1,5 @@
1
  group,benchmark_count,selected_average,best_checkpoint,best_average,selected_rank,selected_margin_vs_best
2
  Core Reasoning Tasks,3,0.719,checkpoints/step_900,0.726,2,-0.007
3
- Language Understanding,4,0.753,checkpoints/step_900,0.755,2,-0.002
4
  Generation Tasks,4,0.686,checkpoints/step_800,0.686,1,0.000
5
- Specialized Capabilities,4,0.774,checkpoints/step_900,0.776,2,-0.002
 
1
  group,benchmark_count,selected_average,best_checkpoint,best_average,selected_rank,selected_margin_vs_best
2
  Core Reasoning Tasks,3,0.719,checkpoints/step_900,0.726,2,-0.007
3
+ Language Understanding,4,0.754,checkpoints/step_900,0.756,2,-0.002
4
  Generation Tasks,4,0.686,checkpoints/step_800,0.686,1,0.000
5
+ Specialized Capabilities,4,0.774,checkpoints/step_900,0.777,2,-0.002
pytorch_model.bin CHANGED
@@ -1,3 +1 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:01934db10231d7b9e1079e726312af2c87b861d0fc64d011a675d6920038397c
3
- size 34
 
1
+ risk-floor-release-card::step_800
 
 
release_evidence.tar CHANGED
Binary files a/release_evidence.tar and b/release_evidence.tar differ
 
release_integrity.json CHANGED
@@ -38,13 +38,13 @@
38
  "figures/fig2.png": "bd81e62dbd42",
39
  "figures/fig3.png": "bd81e62dbd42",
40
  "risk_report.json": "043309ec3016",
41
- "decision_table.csv": "d319cd20296e",
42
- "group_summary.csv": "18fc933f8a23",
43
  "lineage_audit.csv": "972560e8be76",
44
- "release_trace.zip": "d5ffea4dea32",
45
- "release_evidence.tar": "1a78e816f5d7"
46
  },
47
- "payload_digest": "8dc64f57198db9e6",
48
  "payload_digest_line_count": 12,
49
  "payload_digest_algorithm": "sha256_first16(path_tab_size_tab_sha12_lf)",
50
  "manifest_consistency": true
 
38
  "figures/fig2.png": "bd81e62dbd42",
39
  "figures/fig3.png": "bd81e62dbd42",
40
  "risk_report.json": "043309ec3016",
41
+ "decision_table.csv": "1680bb71347f",
42
+ "group_summary.csv": "cd5792dd5dc7",
43
  "lineage_audit.csv": "972560e8be76",
44
+ "release_trace.zip": "d637d034e8bc",
45
+ "release_evidence.tar": "d701f36891fa"
46
  },
47
+ "payload_digest": "6a182c01fdaf95d9",
48
  "payload_digest_line_count": 12,
49
  "payload_digest_algorithm": "sha256_first16(path_tab_size_tab_sha12_lf)",
50
  "manifest_consistency": true
release_trace.zip CHANGED
Binary files a/release_trace.zip and b/release_trace.zip differ
 
upload_manifest.json CHANGED
@@ -31,12 +31,12 @@
31
  "figures/fig2.png": "bd81e62dbd42",
32
  "figures/fig3.png": "bd81e62dbd42",
33
  "risk_report.json": "043309ec3016",
34
- "decision_table.csv": "d319cd20296e",
35
- "group_summary.csv": "18fc933f8a23",
36
  "lineage_audit.csv": "972560e8be76",
37
- "release_trace.zip": "d5ffea4dea32",
38
- "release_evidence.tar": "1a78e816f5d7",
39
- "release_integrity.json": "bb7230b0f956",
40
  "upload_manifest.json": "self-referential"
41
  },
42
  "readback_required": true
 
31
  "figures/fig2.png": "bd81e62dbd42",
32
  "figures/fig3.png": "bd81e62dbd42",
33
  "risk_report.json": "043309ec3016",
34
+ "decision_table.csv": "1680bb71347f",
35
+ "group_summary.csv": "cd5792dd5dc7",
36
  "lineage_audit.csv": "972560e8be76",
37
+ "release_trace.zip": "d637d034e8bc",
38
+ "release_evidence.tar": "d701f36891fa",
39
+ "release_integrity.json": "fb12f1511d57",
40
  "upload_manifest.json": "self-referential"
41
  },
42
  "readback_required": true