SlayerLab
/

Slayer1B-1.09

@@ -2,8 +2,8 @@
   "schema_version": 1,
   "measurement_owner": "SlayerLab / Fabryka AI",
   "status": "descriptive; training-lineage contamination scan pending",
-  "source_aggregate_sha256": "ce0822a324f805bb61f28cc8ea3197e6a9abd97a4591239349dc60141f5289db",
-  "source_generator_sha256": "b792b5a7b4f5218acff4b3d5e9fb63c6f11d39fa26cca6fbd467758c90555a34",
   "benchmark_community_report": "https://github.com/slayerlabs/benchmarks-fabryka-ai/blob/8b547007526971bd2fc35dfe56ff2c7d9b802568/research/results/full-run-2026-11/README.md",
   "community_report_is_separate_measurement": true,
   "protocol": {
@@ -27,14 +27,25 @@
       "mmlu_pl": "acc"
     },
     "mmlu_aggregation": "micro-average over questions",
-    "polish_task_definitions": "OpenGPTX ogx_hellaswagx_pl and ogx_mmlux_pl; source revision 0018eb9b"
   },
   "ci": {
     "level": 0.95,
     "method": "percentile bootstrap via Binomial(n,p)/n for binary item scores",
     "replicates": 10000,
     "base_seed": 20261110,
-    "seed_scope": "base seed plus model label, as source generator",
     "assumption": "items treated as independent; no subject, template or source clusters",
     "interpretation": "pointwise interval for one result; not a paired interval for model differences; not training-seed variance"
   },
@@ -800,203 +811,966 @@
           "correct_primary": 3384
         }
       }
-    }
-  },
-  "additional_measurements": {
-    "benchteam_gsm8k_sst2_20261111": {
-      "measurement_owner": "SlayerLab / Fabryka AI",
-      "status": "descriptive; training-lineage contamination scan pending",
-      "measured_utc_date": "2026-11-11",
-      "source_manifest_sha256": "229c86053bda604716282e314e584f2eac24d870dc5f99e102328536ecd9d95e",
-      "protocol": {
-        "source_repository": "https://github.com/slayerlabs/benchmarks-fabryka-ai",
-        "inventory_commit": "8b547007526971bd2fc35dfe56ff2c7d9b802568",
-        "source_protocol_id": "bench50-v0.3.1-fullB",
-        "scope": "two-task subset, full evaluated splits; not a full 50-task result",
-        "model_adapter": "slayer",
-        "lm_eval_version": "0.4.9.1",
-        "dtype": "float32",
-        "batch_size": 1,
-        "context_limit": 4096,
-        "task_settings": {
-          "gsm8k": {
-            "task_id": "T33",
-            "split": "test",
-            "few_shot": 5,
-            "metric": "exact_match,strict-match"
           },
-          "sst2": {
-            "task_id": "T21",
-            "split": "validation",
-            "few_shot": 0,
-            "metric": "acc,none"
-          }
         },
-        "metric_units": "fraction; multiply by 100 for card percentages",
-        "ci_interpretation": "pointwise intervals for individual scores as reported; not confidence intervals for model differences or training-seed variance",
-        "dataset_identity": "source_id_digest pins the evaluated item set; dataset revisions are not supplied in these aggregate reports"
-      },
-      "models": {
-        "1.07": {
-          "step": 22327,
-          "weights_published": true,
-          "repo": "SlayerLab/Slayer1B-1.07",
-          "revision": "a090b54e658fb57dc5166c1fcc90cc6bf5c80106",
-          "weights_sha256": "d826b4a11bf4eaccee10fe43ec19a4f6d515dfc5f3e5c859adb631f1a0b76ad5",
-          "source_run_id": "niezalezny-SlayerLab_Slayer1B-1.07-a090b54e-20261111T012858Z",
-          "source_aggregate_sha256": "4639f14fc11b2c254f91fde7b7f606cd90a8480a443df37c6671efb3c09c8c7b",
-          "tasks": {
-            "gsm8k": {
-              "task_id": "T33",
-              "metric": "exact_match,strict-match",
-              "value": 0.015163002274450341,
-              "unit": "score",
-              "ci95": {
-                "lower": 0.009097801364670205,
-                "upper": 0.021986353297952996,
-                "level": 0.95,
-                "method": "paired_bootstrap",
-                "replicates": 10000,
-                "seed": 20261028
-              },
-              "n_scored": 1319,
-              "n_population": 1319,
-              "contamination": "unscanned",
-              "state": "ok",
-              "runner_sha256": "c906036b4e597e2c17ad7f1f6441e72fbf8e066946c2a7731966612674e3995e",
-              "results_sha256": "d20e0d2b1f0e21915f857c6c4218f43f4c1db4c2799e28a78c3618d4fcfcf1b6",
-              "source_id_digest": "a32d2c4b15b4e22c87fc4e4717cf476baa2aa90a17f61230cef747c0a3bb61d0"
-            },
-            "sst2": {
-              "task_id": "T21",
-              "metric": "acc,none",
-              "value": 0.5103211009174312,
-              "unit": "score",
-              "ci95": {
-                "lower": 0.47706422018348627,
-                "upper": 0.5435779816513762,
-                "level": 0.95,
-                "method": "paired_bootstrap",
-                "replicates": 10000,
-                "seed": 20261028
-              },
-              "n_scored": 872,
-              "n_population": 872,
-              "contamination": "unscanned",
-              "state": "ok",
-              "runner_sha256": "c906036b4e597e2c17ad7f1f6441e72fbf8e066946c2a7731966612674e3995e",
-              "results_sha256": "f1965fb1a3651c6ab526a82d9bc35504b8c4f92c4789b471d61fa6a37e54c83f",
-              "source_id_digest": "7a716d60452e40d779aaad8da40aa11cdfb66cf0a069b6f0a5c5b38c78819ce7"
-            }
-          }
         },
-        "1.08": {
-          "step": 1709,
-          "weights_published": true,
-          "repo": "SlayerLab/Slayer1B-1.08",
-          "revision": "5f854a0cc7160594d54d1788d37a44e8ebabe16c",
-          "weights_sha256": "1c510d9a59d3b9e62360d1db57e08fa450fb9b09df849df7f2eef132ea6715a4",
-          "source_run_id": "niezalezny-SlayerLab_Slayer1B-1.08-5f854a0c-20261111T023631Z",
-          "source_aggregate_sha256": "3ee230fb5029cf9c6424c27bae2b60dfd345998b2bee42021cb53e625021615b",
-          "tasks": {
-            "gsm8k": {
-              "task_id": "T33",
-              "metric": "exact_match,strict-match",
-              "value": 0.009097801364670205,
-              "unit": "score",
-              "ci95": {
-                "lower": 0.004548900682335102,
-                "upper": 0.014404852160727824,
-                "level": 0.95,
-                "method": "paired_bootstrap",
-                "replicates": 10000,
-                "seed": 20261028
-              },
-              "n_scored": 1319,
-              "n_population": 1319,
-              "contamination": "unscanned",
-              "state": "ok",
-              "runner_sha256": "c906036b4e597e2c17ad7f1f6441e72fbf8e066946c2a7731966612674e3995e",
-              "results_sha256": "7e2f64e0958f9e5897debe2804cfb25c804b4abe1b8bdd440f7885d16b98f3e6",
-              "source_id_digest": "a32d2c4b15b4e22c87fc4e4717cf476baa2aa90a17f61230cef747c0a3bb61d0"
-            },
-            "sst2": {
-              "task_id": "T21",
-              "metric": "acc,none",
-              "value": 0.5045871559633027,
-              "unit": "score",
-              "ci95": {
-                "lower": 0.4713302752293578,
-                "upper": 0.536697247706422,
-                "level": 0.95,
-                "method": "paired_bootstrap",
-                "replicates": 10000,
-                "seed": 20261028
-              },
-              "n_scored": 872,
-              "n_population": 872,
-              "contamination": "unscanned",
-              "state": "ok",
-              "runner_sha256": "c906036b4e597e2c17ad7f1f6441e72fbf8e066946c2a7731966612674e3995e",
-              "results_sha256": "1e1a7b5d6c41dff7140208736213f308a5be62986bef1e9bf5b7aeb3e176a48a",
-              "source_id_digest": "7a716d60452e40d779aaad8da40aa11cdfb66cf0a069b6f0a5c5b38c78819ce7"
-            }
-          }
         },
-        "1.09-k10000": {
-          "step": 10000,
-          "weights_published": false,
-          "repo": "Slayer1B-1.09-k10000-lokalny",
-          "revision": null,
-          "weights_sha256": "12939d84ef0d06b350773fb30837390306d899ded315d1721b52c7615a87317a",
-          "source_run_id": "niezalezny-Slayer1B-1.09-k10000-lokalny-lokalny-20261111T025334Z",
-          "source_aggregate_sha256": "042b2639bc325aeb7fb60787bc8205d1384557e37dbb7b01ba5489db41ebdfe8",
-          "tasks": {
-            "gsm8k": {
-              "task_id": "T33",
-              "metric": "exact_match,strict-match",
-              "value": 0.009855951478392721,
-              "unit": "score",
-              "ci95": {
-                "lower": 0.004548900682335102,
-                "upper": 0.015163002274450341,
-                "level": 0.95,
-                "method": "paired_bootstrap",
-                "replicates": 10000,
-                "seed": 20261028
-              },
-              "n_scored": 1319,
-              "n_population": 1319,
-              "contamination": "unscanned",
-              "state": "ok",
-              "runner_sha256": "c906036b4e597e2c17ad7f1f6441e72fbf8e066946c2a7731966612674e3995e",
-              "results_sha256": "5cf7805fa9c7d3d65d6af44920b6659287c621294efe4d7cde62f2055e65562e",
-              "source_id_digest": "a32d2c4b15b4e22c87fc4e4717cf476baa2aa90a17f61230cef747c0a3bb61d0"
-            },
-            "sst2": {
-              "task_id": "T21",
-              "metric": "acc,none",
-              "value": 0.5825688073394495,
-              "unit": "score",
-              "ci95": {
-                "lower": 0.5493119266055045,
-                "upper": 0.6135321100917431,
-                "level": 0.95,
-                "method": "paired_bootstrap",
-                "replicates": 10000,
-                "seed": 20261028
-              },
-              "n_scored": 872,
-              "n_population": 872,
-              "contamination": "unscanned",
-              "state": "ok",
-              "runner_sha256": "c906036b4e597e2c17ad7f1f6441e72fbf8e066946c2a7731966612674e3995e",
-              "results_sha256": "d47706c62088fe772f1c26d2ed13f52dfecc8d832bbdecd14dd24c89086e9659",
-              "source_id_digest": "7a716d60452e40d779aaad8da40aa11cdfb66cf0a069b6f0a5c5b38c78819ce7"
-            }
-          }
-        }
-      }
-    }
-  }
 }

   "schema_version": 1,
   "measurement_owner": "SlayerLab / Fabryka AI",
   "status": "descriptive; training-lineage contamination scan pending",
+  "source_aggregate_sha256": "fc6f67a20bdc6ada687349d35de26f6811e27766dfe1aa197082000c03535796",
+  "source_generator_sha256": null,
   "benchmark_community_report": "https://github.com/slayerlabs/benchmarks-fabryka-ai/blob/8b547007526971bd2fc35dfe56ff2c7d9b802568/research/results/full-run-2026-11/README.md",
   "community_report_is_separate_measurement": true,
   "protocol": {
       "mmlu_pl": "acc"
     },
     "mmlu_aggregation": "micro-average over questions",
+    "polish_task_definitions": "OpenGPTX ogx_hellaswagx_pl and ogx_mmlux_pl; source revision 0018eb9b",
+    "task_protocols": {
+      "arc_challenge": "lm-eval arc_challenge: ctx='Question: {question}\\nAnswer:', kont=' '+choices.text[i], gold=choices.label.index(answerKey); 0-shot; LL = suma log p tokenów kontynuacji; start id 0; add_special_tokens=False; _encode_pair jak HFLM; obcięcie z lewej do block; fp32",
+      "hellaswag": "lm-eval hellaswag: process_docs (preprocess(activity_label+': '+ctx_a+' '+ctx_b.capitalize()), endings preprocess), ctx=query, kont=' '+ending, gold=int(label); 0-shot; LL = suma log p tokenów kontynuacji; start id 0; add_special_tokens=False; _encode_pair jak HFLM; obcięcie z lewej do block; fp32",
+      "piqa": "lm-eval piqa (baber/piqa): ctx='Question: {goal}\\nAnswer:', kont=' '+[sol1,sol2][i], gold=int(label); 0-shot; LL = suma log p tokenów kontynuacji; start id 0; add_special_tokens=False; _encode_pair jak HFLM; obcięcie z lewej do block; fp32",
+      "winogrande": "lm-eval winogrande (partial scoring): ctx=sentence[:idx('_')]+option_i, kont=' '+sentence[idx+1:].strip() (wspólna), gold={'1':0,'2':1}[answer]; 0-shot; LL = suma log p tokenów kontynuacji; start id 0; add_special_tokens=False; _encode_pair jak HFLM; obcięcie z lewej do block; fp32",
+      "boolq": "lm-eval boolq v2: ctx='{passage}\\nQuestion: {question}?\\nAnswer:', kont=' no'|' yes', gold=int(label); 0-shot; LL = suma log p tokenów kontynuacji; start id 0; add_special_tokens=False; _encode_pair jak HFLM; obcięcie z lewej do block; fp32",
+      "openbookqa": "lm-eval openbookqa: ctx=question_stem, kont=' '+choices.text[i], gold=choices.label.index(answerKey.lstrip()); 0-shot; LL = suma log p tokenów kontynuacji; start id 0; add_special_tokens=False; _encode_pair jak HFLM; obcięcie z lewej do block; fp32",
+      "belebele_pol": "lm-eval belebele_pol_Latn 0-shot: ctx='P: {flores_passage}\\nQ: {question.strip()}\\nA: {mc_answer1}\\nB: ...\\nC: ...\\nD: {mc_answer4}\\nAnswer:', kont=' A'..' D', gold=int(correct_answer_num)-1; 0-shot; LL = suma log p tokenów kontynuacji; start id 0; add_special_tokens=False; _encode_pair jak HFLM; obcięcie z lewej do block; fp32",
+      "hellaswag_pl": "OpenGPTX ogx_hellaswagx_pl @0018eb9b (= lm-eval hellaswag process_docs na polskich polach); 0-shot; LL = suma log p tokenów kontynuacji; start id 0; add_special_tokens=False; _encode_pair jak HFLM; obcięcie z lewej do block; fp32",
+      "mmlu_pl": "OpenGPTX ogx_mmlux_pl-* @0018eb9b 0-shot: ctx=description(PL, doklejone BEZ separatora, jak w forku)+'{question.strip()}\\nA. {c0}\\nB. {c1}\\nC. {c2}\\nD. {c3}\\nOdpowiedź:', kont=' A'..' D', gold=answer; acc = mikro po 14042 pozycjach (jak grupa lm-eval mmlu, wagi = rozmiary); 0-shot; LL = suma log p tokenów kontynuacji; start id 0; add_special_tokens=False; _encode_pair jak HFLM; obcięcie z lewej do block; fp32"
+    }
   },
   "ci": {
     "level": 0.95,
     "method": "percentile bootstrap via Binomial(n,p)/n for binary item scores",
     "replicates": 10000,
     "base_seed": 20261110,
+    "seed_scope": "Updated aggregate records seed 20261110; stream derivation is not supplied. Previous aggregate documented a model-specific stream.",
     "assumption": "items treated as independent; no subject, template or source clusters",
     "interpretation": "pointwise interval for one result; not a paired interval for model differences; not training-seed variance"
   },
           "correct_primary": 3384
         }
       }
+    },
+    "1.09-k10000": {
+      "step": 10000,
+      "weights_sha256": "12939d84ef0d06b350773fb30837390306d899ded315d1721b52c7615a87317a",
+      "weights_published": false,
+      "source_results_sha256": "3fff987f3b9d74eaba1e121a6a769247fbc12c79043bd5cb5d442c8dafeb9a69",
+      "evaluator_sha256": "ae6658f9c83a484e373dc6d14476a7aeb98a905c27cd6d97ea234fb56140e016",
+      "tasks": {
+        "arc_challenge": {
+          "n": 1172,
+          "n_pelne": 1172,
+          "dataset_repo": "allenai/ai2_arc",
+          "dataset_sha": "210d026faf9955653af8916fad021475a3f00453",
+          "pliki_sha256": {
+            "ARC-Challenge/test-00000-of-00001.parquet": "62f03257e737aed263f55c6abf87c7bb0028a44a6bdd2a26eb1279eb42c1d1e9"
           },
+          "acc": 0.292662,
+          "acc_ci95": [
+            0.267044,
+            0.319113
+          ],
+          "acc_norm": 0.304608,
+          "acc_norm_ci95": [
+            0.278157,
+            0.331058
+          ],
+          "split": "test",
+          "correct_primary": 357
         },
+        "hellaswag": {
+          "n": 10042,
+          "n_pelne": 10042,
+          "dataset_repo": "Rowan/hellaswag",
+          "dataset_sha": "218ec52e09a7e7462a5400043bb9a69a41d06b76",
+          "pliki_sha256": {
+            "data/validation-00000-of-00001.parquet": "899813071e1e95efafec90f856e1987d2150fa4d020fc005df6962c259f660cd"
+          },
+          "acc": 0.386178,
+          "acc_ci95": [
+            0.376817,
+            0.39574
+          ],
+          "acc_norm": 0.489345,
+          "acc_norm_ci95": [
+            0.479685,
+            0.499104
+          ],
+          "split": "validation",
+          "correct_primary": 4914
         },
+        "piqa": {
+          "n": 1838,
+          "n_pelne": 1838,
+          "dataset_repo": "baber/piqa",
+          "dataset_sha": "142f6d7367fd9877f0fb3b5734ea6a545f54cdd1",
+          "pliki_sha256": {
+            "piqa_validation.parquet": "2b17ee1215b1ceb2ddabf4c90a05c5a8b833bef9d207960f57d111c43a2080f3"
+          },
+          "acc": 0.696953,
+          "acc_ci95": [
+            0.675734,
+            0.717628
+          ],
+          "acc_norm": 0.68988,
+          "acc_norm_ci95": [
+            0.669206,
+            0.711099
+          ],
+          "split": "validation",
+          "correct_primary": 1268
         },
+        "winogrande": {
+          "n": 1267,
+          "n_pelne": 1267,
+          "dataset_repo": "allenai/winogrande",
+          "dataset_sha": "01e74176c63542e6b0bcb004dcdea22d94fb67b5",
+          "pliki_sha256": {
+            "winogrande_xl/validation-00000-of-00001.parquet": "f9d914d1818c0aba98cabdf8af2bc4bd943c462b9c5e7062647e05c2ce3315d1"
+          },
+          "acc": 0.549329,
+          "acc_ci95": [
+            0.521705,
+            0.576164
+          ],
+          "acc_norm": 0.549329,
+          "acc_norm_ci95": [
+            0.521705,
+            0.576953
+          ],
+          "split": "validation",
+          "correct_primary": 696
+        },
+        "boolq": {
+          "n": 3270,
+          "n_pelne": 3270,
+          "dataset_repo": "aps/super_glue",
+          "dataset_sha": "3de24cf8022e94f4ee4b9d55a6f539891524d646",
+          "pliki_sha256": {
+            "boolq/validation-00000-of-00001.parquet": "f46438b8cce048184552a2bfcdb8252eadb3a2602f0c22334caccd9b6ea547d6"
+          },
+          "acc": 0.611315,
+          "acc_ci95": [
+            0.594495,
+            0.627829
+          ],
+          "acc_norm": 0.62263,
+          "acc_norm_ci95": [
+            0.60581,
+            0.639144
+          ],
+          "split": "validation",
+          "correct_primary": 1999
+        },
+        "openbookqa": {
+          "n": 500,
+          "n_pelne": 500,
+          "dataset_repo": "allenai/openbookqa",
+          "dataset_sha": "388097ea7776314e93a529163e0fea805b8a6454",
+          "pliki_sha256": {
+            "main/test-00000-of-00001.parquet": "cd5483e366daa230c1c87bbdc512d8b7229f14f6dd04d19fc8b1a3855aaaa8a3"
+          },
+          "acc": 0.242,
+          "acc_ci95": [
+            0.206,
+            0.28
+          ],
+          "acc_norm": 0.344,
+          "acc_norm_ci95": [
+            0.302,
+            0.386
+          ],
+          "split": "test",
+          "correct_primary": 172
+        },
+        "belebele_pol": {
+          "n": 900,
+          "n_pelne": 900,
+          "dataset_repo": "facebook/belebele",
+          "dataset_sha": "7899cdfa4e1e0d733fd77c848e2c273cb1d32be2",
+          "pliki_sha256": {
+            "data/pol_Latn.jsonl": "1699bb60b65e258d71989c541027c5439eb36a4f5032a5be392137bf5a1b6bec"
+          },
+          "acc": 0.241111,
+          "acc_ci95": [
+            0.213333,
+            0.27
+          ],
+          "acc_norm": 0.241111,
+          "acc_norm_ci95": [
+            0.213333,
+            0.27
+          ],
+          "split": "pol_Latn evaluation set",
+          "correct_primary": 217
+        },
+        "hellaswag_pl": {
+          "n": 10042,
+          "n_pelne": 10042,
+          "dataset_repo": "Eurolingua/hellaswagx",
+          "dataset_sha": "c38911a204759866dbcde35eeb584c36caf11a8b",
+          "pliki_sha256": {
+            "hellaswag_PL_validation.jsonl": "acdfc6d397885a4c86de178a6e5f9b7e9b90a7ff81ffcd6e5e0d3d0788dd3a16"
+          },
+          "acc": 0.354013,
+          "acc_ci95": [
+            0.344652,
+            0.363274
+          ],
+          "acc_norm": 0.455089,
+          "acc_norm_ci95": [
+            0.44513,
+            0.464848
+          ],
+          "split": "validation",
+          "correct_primary": 4570
+        },
+        "mmlu_pl": {
+          "n": 14042,
+          "n_pelne": 14042,
+          "dataset_repo": "Eurolingua/mmlux",
+          "dataset_sha": "7fb62cf05c1ce9ea550dbfd2ea6c145204344882",
+          "pliki_sha256": {
+            "hendrycks_abstract_algebra_PL_test.jsonl": "4905c6ef8d306c2ff1e1771ef8c15b63152bb88effec11844217893b81fcc3ea",
+            "hendrycks_anatomy_PL_test.jsonl": "5cba9bc54a6f0bdaddcd2e0e23155fccfeab469b561f03cc571e1387abeb9a38",
+            "hendrycks_astronomy_PL_test.jsonl": "00dc5fca7a8be3d8bef9e24ad9018e5417fb07703ebd2d0039c2c82418272b03",
+            "hendrycks_business_ethics_PL_test.jsonl": "9c9542880c720e2d0f5a768ab2d6602bdf5c394d66e51ff352af79aaca6bbd96",
+            "hendrycks_clinical_knowledge_PL_test.jsonl": "79fd5fd7a17dc7c7d84a9e9b138cf13c2dcaf6ab332eb1a0f299a7b52dc89c61",
+            "hendrycks_college_biology_PL_test.jsonl": "b49d766fb2c4160809d3907b846cd6c5c609299f47d9096bcf3208b843392c48",
+            "hendrycks_college_chemistry_PL_test.jsonl": "36c6fd391411789310dca0bbd36aec9aae8784df2bac2e6838e0895eebd97f1d",
+            "hendrycks_college_computer_science_PL_test.jsonl": "90ab29151dc4b8c7e13893516474201b73e8f9765cdf87225b6b9cacdd540e70",
+            "hendrycks_college_mathematics_PL_test.jsonl": "84411ee6464d01f24f2482bccb451635fe75d8169e33932ce6e067c1be3d37ee",
+            "hendrycks_college_medicine_PL_test.jsonl": "faa451d891c8a67b15bae4921c49a8afea448ac7b1b6095646c3570ac3040606",
+            "hendrycks_college_physics_PL_test.jsonl": "bfcaac4cb6d13e7cbcc6ebfb7ea6a9ab0d0ec07613b1b10ca679e632106b4bc2",
+            "hendrycks_computer_security_PL_test.jsonl": "c8324fb94b5b4b7e17440520ea20a22d8c03c3b86da81824059602982463cf8a",
+            "hendrycks_conceptual_physics_PL_test.jsonl": "5227e303a5437f437a4ad9d7eb75917d36051ce635af9f43a92d9525918eac64",
+            "hendrycks_econometrics_PL_test.jsonl": "9bf1d2a2c9eb4f5692b3876919dbc4e679d43d3ea2b64dc8fa2c17703a3e696b",
+            "hendrycks_electrical_engineering_PL_test.jsonl": "a431be1d342cc1eca6aa408688b78751c74621056a4fa4c31e7dfda443e4be6f",
+            "hendrycks_elementary_mathematics_PL_test.jsonl": "9cdb8692fc08d6d8ac5dec24d89f386863b6a02480eb3b8c5be641f15700e667",
+            "hendrycks_formal_logic_PL_test.jsonl": "0f30d48346a0cd2f136463b283d8bd3fe085de67da2a34d90cac689740f9b5be",
+            "hendrycks_global_facts_PL_test.jsonl": "21c753017550f0d31d7c2cdbe82637c877682a9edf8ab913de720e79c82b374f",
+            "hendrycks_high_school_biology_PL_test.jsonl": "7b086933dd97cfc392fba5f78c3523ddc0302cf2b7caed7428a75bfdf9fbb7f2",
+            "hendrycks_high_school_chemistry_PL_test.jsonl": "85f836040875313d60cf9241bd9e827a3cdb770cf159ddf3f7032edb293871ec",
+            "hendrycks_high_school_computer_science_PL_test.jsonl": "034b7b497ada5cc1146c6c5fa02291094d5da512bc0c618ae2b967bd941a70c1",
+            "hendrycks_high_school_european_history_PL_test.jsonl": "023ce8f773cb5dccd86658165a86a5bc1b0d83d53cede5a22c8f7a027582e2a4",
+            "hendrycks_high_school_geography_PL_test.jsonl": "3cce3d6defc5cad9742f9db4407ec5176e6b5d1634b0f8ecb4cfe9f4cabfbe5a",
+            "hendrycks_high_school_government_and_politics_PL_test.jsonl": "9c73c16602cd663acf3b08cdecef6aa8a2c019ed647f02cb970f078d88e3ec7d",
+            "hendrycks_high_school_macroeconomics_PL_test.jsonl": "9425c91f59bd35355262f0c60b04b479a2e700064751efddadb9a28cbc83974d",
+            "hendrycks_high_school_mathematics_PL_test.jsonl": "80c73fa2b882104047260829a57cdca588d1ec9ba690b8a8a5cb87f65283e735",
+            "hendrycks_high_school_microeconomics_PL_test.jsonl": "aebc508914a9b32f5dd1bdc3ebd1375a619791214cb67968ccc6d5712675cbe8",
+            "hendrycks_high_school_physics_PL_test.jsonl": "bba7751511cada0b7b0b07133f2a961e5f046e174e201628a6722a279ab4e47f",
+            "hendrycks_high_school_psychology_PL_test.jsonl": "adf5943b42493ae59f9310028a45adffaace2359d52b8cbb1fa9cc41d6371e45",
+            "hendrycks_high_school_statistics_PL_test.jsonl": "14ffdec0e35d7437d7b2243c3bcbf606843df9ac4eaa044cae95faa3f6e41fdf",
+            "hendrycks_high_school_us_history_PL_test.jsonl": "d8a7946dbf1860b1e99ff0d1948ede18500cc06984ef68786ffed6a8160731f7",
+            "hendrycks_high_school_world_history_PL_test.jsonl": "ae943acd33cf4b23fd3d70d4cc5b0d16f073574c547de91bfc157d552bf63ec0",
+            "hendrycks_human_aging_PL_test.jsonl": "21df1fd8d194a6e453e71e1583626fc723f1c9b44cc571ef7d6074450754ed82",
+            "hendrycks_human_sexuality_PL_test.jsonl": "d7c467fcee3671069986948b69e6ec69c2630173207025ce4925719bb1a880d4",
+            "hendrycks_international_law_PL_test.jsonl": "f0b6f86e54390b63fda0ecf07c7c55f87d30864f6a698c9819f8d8e6eca4d746",
+            "hendrycks_jurisprudence_PL_test.jsonl": "9d7f42f41510da5d835d8df834c4bf6bcd2e03e658cc906ec334810b30ebbe96",
+            "hendrycks_logical_fallacies_PL_test.jsonl": "8b7cd34e9126adcfbc4b9b7bee8345d97352ec892221a0e98fa268210b04ee51",
+            "hendrycks_machine_learning_PL_test.jsonl": "f654c60a51877a9d269471ab01b54b0d4d20f9cc07f7fb57512fb3e90d5a8608",
+            "hendrycks_management_PL_test.jsonl": "628374b8f9beabefe43c2b1dda2c07760eb455d2e15f9e58f58506d61f0fc3bb",
+            "hendrycks_marketing_PL_test.jsonl": "31dc1eef1527c98642080c0b3866ea113bda77c2a03c5155aafe382206ffacdb",
+            "hendrycks_medical_genetics_PL_test.jsonl": "f05fcc33d029460a7f7916ec83debfc746ced5ea931f7a56b63fe1245819e4a4",
+            "hendrycks_miscellaneous_PL_test.jsonl": "94be7854c05aef25ebca4ae1dc562f37172b5346a14fb760b8e89eefea588809",
+            "hendrycks_moral_disputes_PL_test.jsonl": "245ad1cf635274e5ac81559542cb21638b7c729e8a811f76271f4301937d6127",
+            "hendrycks_moral_scenarios_PL_test.jsonl": "d60b08bb71451594385db5ff6ffdef4ca11fec801dc86d681e68f1727a690f9b",
+            "hendrycks_nutrition_PL_test.jsonl": "3cfe618d8e23e3c8e5c752544f07a71b90b4c84d4bc5d673026e8be4de4bf285",
+            "hendrycks_philosophy_PL_test.jsonl": "a1939f9773085bdae7e127fce099afc4153ec4cbe5f66e126f12599feaf76639",
+            "hendrycks_prehistory_PL_test.jsonl": "42eaf871b859f22321dfdec21b6b6b4a3c8f1435c2d4cd2da46dc9fb5ceced88",
+            "hendrycks_professional_accounting_PL_test.jsonl": "1fb6440616a195b94d3de760af5cacd05656667dfac595e2601504cda407c25e",
+            "hendrycks_professional_law_PL_test.jsonl": "fb86b175ecdfe8b56d4991bf6d6705a58f1d89a3efda2a8eb5f01f8082b21e57",
+            "hendrycks_professional_medicine_PL_test.jsonl": "ec818eb13a66e631ab82a10bae1536377dcd4bcadf2321c964b7718c60992080",
+            "hendrycks_professional_psychology_PL_test.jsonl": "72ddab1aaddc26266f4964941334a84d9243f63c329d2d9349bdcaa20efa6bc3",
+            "hendrycks_public_relations_PL_test.jsonl": "2e16835d88ba1ebb6f70a526bd85b9830cb1d3e14e2358a43ce15a3d548232e1",
+            "hendrycks_security_studies_PL_test.jsonl": "c476a634f4f48ff025869158057de726a8080bee253d88921d90252204aa3e4c",
+            "hendrycks_sociology_PL_test.jsonl": "dd56860da95ed793943c9c201f9aa8f5aa8d83753b1bbcbc440fff7ac30f881a",
+            "hendrycks_us_foreign_policy_PL_test.jsonl": "95183b3fcd02d17f5bb7d359b709b3fde13bf6f0f0f1375ab676b0e7dda36786",
+            "hendrycks_virology_PL_test.jsonl": "99e2c6f80907315317619f8c92f8d501ae484619c1169e36d9ad8fa59abfbda7",
+            "hendrycks_world_religions_PL_test.jsonl": "4631dbc58d379cbc46751515f36af682f8bf53db36f3345dbe090d5a00b7b24a"
+          },
+          "acc": 0.233229,
+          "acc_ci95": [
+            0.22625,
+            0.240279
+          ],
+          "acc_norm": 0.233229,
+          "acc_norm_ci95": [
+            0.22625,
+            0.240422
+          ],
+          "split": "test; micro-average over subjects",
+          "correct_primary": 3275
+        }
+      }
+    },
+    "1.09-k15000": {
+      "step": 15000,
+      "weights_sha256": "d5e3927cc5bfe380f6acb9a2de089ef776ae61a57a70d5a719da9f4da3932f9d",
+      "weights_published": false,
+      "source_results_sha256": "1bbb4ce856f45ed4ebacc1911431b749f6df4ed8f5caf292437c9424f03959bd",
+      "evaluator_sha256": "ae6658f9c83a484e373dc6d14476a7aeb98a905c27cd6d97ea234fb56140e016",
+      "tasks": {
+        "arc_challenge": {
+          "n": 1172,
+          "n_pelne": 1172,
+          "dataset_repo": "allenai/ai2_arc",
+          "dataset_sha": "210d026faf9955653af8916fad021475a3f00453",
+          "pliki_sha256": {
+            "ARC-Challenge/test-00000-of-00001.parquet": "62f03257e737aed263f55c6abf87c7bb0028a44a6bdd2a26eb1279eb42c1d1e9"
+          },
+          "acc": 0.309727,
+          "acc_ci95": [
+            0.283276,
+            0.337031
+          ],
+          "acc_norm": 0.325085,
+          "acc_norm_ci95": [
+            0.298635,
+            0.352389
+          ],
+          "split": "test",
+          "correct_primary": 381
+        },
+        "hellaswag": {
+          "n": 10042,
+          "n_pelne": 10042,
+          "dataset_repo": "Rowan/hellaswag",
+          "dataset_sha": "218ec52e09a7e7462a5400043bb9a69a41d06b76",
+          "pliki_sha256": {
+            "data/validation-00000-of-00001.parquet": "899813071e1e95efafec90f856e1987d2150fa4d020fc005df6962c259f660cd"
+          },
+          "acc": 0.387771,
+          "acc_ci95": [
+            0.378411,
+            0.39753
+          ],
+          "acc_norm": 0.49771,
+          "acc_norm_ci95": [
+            0.487951,
+            0.507568
+          ],
+          "split": "validation",
+          "correct_primary": 4998
+        },
+        "piqa": {
+          "n": 1838,
+          "n_pelne": 1838,
+          "dataset_repo": "baber/piqa",
+          "dataset_sha": "142f6d7367fd9877f0fb3b5734ea6a545f54cdd1",
+          "pliki_sha256": {
+            "piqa_validation.parquet": "2b17ee1215b1ceb2ddabf4c90a05c5a8b833bef9d207960f57d111c43a2080f3"
+          },
+          "acc": 0.695321,
+          "acc_ci95": [
+            0.674646,
+            0.71654
+          ],
+          "acc_norm": 0.68988,
+          "acc_norm_ci95": [
+            0.668662,
+            0.710555
+          ],
+          "split": "validation",
+          "correct_primary": 1268
+        },
+        "winogrande": {
+          "n": 1267,
+          "n_pelne": 1267,
+          "dataset_repo": "allenai/winogrande",
+          "dataset_sha": "01e74176c63542e6b0bcb004dcdea22d94fb67b5",
+          "pliki_sha256": {
+            "winogrande_xl/validation-00000-of-00001.parquet": "f9d914d1818c0aba98cabdf8af2bc4bd943c462b9c5e7062647e05c2ce3315d1"
+          },
+          "acc": 0.567482,
+          "acc_ci95": [
+            0.540647,
+            0.594317
+          ],
+          "acc_norm": 0.567482,
+          "acc_norm_ci95": [
+            0.539858,
+            0.594317
+          ],
+          "split": "validation",
+          "correct_primary": 719
+        },
+        "boolq": {
+          "n": 3270,
+          "n_pelne": 3270,
+          "dataset_repo": "aps/super_glue",
+          "dataset_sha": "3de24cf8022e94f4ee4b9d55a6f539891524d646",
+          "pliki_sha256": {
+            "boolq/validation-00000-of-00001.parquet": "f46438b8cce048184552a2bfcdb8252eadb3a2602f0c22334caccd9b6ea547d6"
+          },
+          "acc": 0.598777,
+          "acc_ci95": [
+            0.581957,
+            0.615291
+          ],
+          "acc_norm": 0.619572,
+          "acc_norm_ci95": [
+            0.603058,
+            0.636697
+          ],
+          "split": "validation",
+          "correct_primary": 1958
+        },
+        "openbookqa": {
+          "n": 500,
+          "n_pelne": 500,
+          "dataset_repo": "allenai/openbookqa",
+          "dataset_sha": "388097ea7776314e93a529163e0fea805b8a6454",
+          "pliki_sha256": {
+            "main/test-00000-of-00001.parquet": "cd5483e366daa230c1c87bbdc512d8b7229f14f6dd04d19fc8b1a3855aaaa8a3"
+          },
+          "acc": 0.252,
+          "acc_ci95": [
+            0.214,
+            0.29
+          ],
+          "acc_norm": 0.36,
+          "acc_norm_ci95": [
+            0.318,
+            0.402
+          ],
+          "split": "test",
+          "correct_primary": 180
+        },
+        "belebele_pol": {
+          "n": 900,
+          "n_pelne": 900,
+          "dataset_repo": "facebook/belebele",
+          "dataset_sha": "7899cdfa4e1e0d733fd77c848e2c273cb1d32be2",
+          "pliki_sha256": {
+            "data/pol_Latn.jsonl": "1699bb60b65e258d71989c541027c5439eb36a4f5032a5be392137bf5a1b6bec"
+          },
+          "acc": 0.225556,
+          "acc_ci95": [
+            0.198889,
+            0.253333
+          ],
+          "acc_norm": 0.225556,
+          "acc_norm_ci95": [
+            0.197778,
+            0.253333
+          ],
+          "split": "pol_Latn evaluation set",
+          "correct_primary": 203
+        },
+        "hellaswag_pl": {
+          "n": 10042,
+          "n_pelne": 10042,
+          "dataset_repo": "Eurolingua/hellaswagx",
+          "dataset_sha": "c38911a204759866dbcde35eeb584c36caf11a8b",
+          "pliki_sha256": {
+            "hellaswag_PL_validation.jsonl": "acdfc6d397885a4c86de178a6e5f9b7e9b90a7ff81ffcd6e5e0d3d0788dd3a16"
+          },
+          "acc": 0.34286,
+          "acc_ci95": [
+            0.333599,
+            0.352121
+          ],
+          "acc_norm": 0.441844,
+          "acc_norm_ci95": [
+            0.432085,
+            0.451603
+          ],
+          "split": "validation",
+          "correct_primary": 4437
+        },
+        "mmlu_pl": {
+          "n": 14042,
+          "n_pelne": 14042,
+          "dataset_repo": "Eurolingua/mmlux",
+          "dataset_sha": "7fb62cf05c1ce9ea550dbfd2ea6c145204344882",
+          "pliki_sha256": {
+            "hendrycks_abstract_algebra_PL_test.jsonl": "4905c6ef8d306c2ff1e1771ef8c15b63152bb88effec11844217893b81fcc3ea",
+            "hendrycks_anatomy_PL_test.jsonl": "5cba9bc54a6f0bdaddcd2e0e23155fccfeab469b561f03cc571e1387abeb9a38",
+            "hendrycks_astronomy_PL_test.jsonl": "00dc5fca7a8be3d8bef9e24ad9018e5417fb07703ebd2d0039c2c82418272b03",
+            "hendrycks_business_ethics_PL_test.jsonl": "9c9542880c720e2d0f5a768ab2d6602bdf5c394d66e51ff352af79aaca6bbd96",
+            "hendrycks_clinical_knowledge_PL_test.jsonl": "79fd5fd7a17dc7c7d84a9e9b138cf13c2dcaf6ab332eb1a0f299a7b52dc89c61",
+            "hendrycks_college_biology_PL_test.jsonl": "b49d766fb2c4160809d3907b846cd6c5c609299f47d9096bcf3208b843392c48",
+            "hendrycks_college_chemistry_PL_test.jsonl": "36c6fd391411789310dca0bbd36aec9aae8784df2bac2e6838e0895eebd97f1d",
+            "hendrycks_college_computer_science_PL_test.jsonl": "90ab29151dc4b8c7e13893516474201b73e8f9765cdf87225b6b9cacdd540e70",
+            "hendrycks_college_mathematics_PL_test.jsonl": "84411ee6464d01f24f2482bccb451635fe75d8169e33932ce6e067c1be3d37ee",
+            "hendrycks_college_medicine_PL_test.jsonl": "faa451d891c8a67b15bae4921c49a8afea448ac7b1b6095646c3570ac3040606",
+            "hendrycks_college_physics_PL_test.jsonl": "bfcaac4cb6d13e7cbcc6ebfb7ea6a9ab0d0ec07613b1b10ca679e632106b4bc2",
+            "hendrycks_computer_security_PL_test.jsonl": "c8324fb94b5b4b7e17440520ea20a22d8c03c3b86da81824059602982463cf8a",
+            "hendrycks_conceptual_physics_PL_test.jsonl": "5227e303a5437f437a4ad9d7eb75917d36051ce635af9f43a92d9525918eac64",
+            "hendrycks_econometrics_PL_test.jsonl": "9bf1d2a2c9eb4f5692b3876919dbc4e679d43d3ea2b64dc8fa2c17703a3e696b",
+            "hendrycks_electrical_engineering_PL_test.jsonl": "a431be1d342cc1eca6aa408688b78751c74621056a4fa4c31e7dfda443e4be6f",
+            "hendrycks_elementary_mathematics_PL_test.jsonl": "9cdb8692fc08d6d8ac5dec24d89f386863b6a02480eb3b8c5be641f15700e667",
+            "hendrycks_formal_logic_PL_test.jsonl": "0f30d48346a0cd2f136463b283d8bd3fe085de67da2a34d90cac689740f9b5be",
+            "hendrycks_global_facts_PL_test.jsonl": "21c753017550f0d31d7c2cdbe82637c877682a9edf8ab913de720e79c82b374f",
+            "hendrycks_high_school_biology_PL_test.jsonl": "7b086933dd97cfc392fba5f78c3523ddc0302cf2b7caed7428a75bfdf9fbb7f2",
+            "hendrycks_high_school_chemistry_PL_test.jsonl": "85f836040875313d60cf9241bd9e827a3cdb770cf159ddf3f7032edb293871ec",
+            "hendrycks_high_school_computer_science_PL_test.jsonl": "034b7b497ada5cc1146c6c5fa02291094d5da512bc0c618ae2b967bd941a70c1",
+            "hendrycks_high_school_european_history_PL_test.jsonl": "023ce8f773cb5dccd86658165a86a5bc1b0d83d53cede5a22c8f7a027582e2a4",
+            "hendrycks_high_school_geography_PL_test.jsonl": "3cce3d6defc5cad9742f9db4407ec5176e6b5d1634b0f8ecb4cfe9f4cabfbe5a",
+            "hendrycks_high_school_government_and_politics_PL_test.jsonl": "9c73c16602cd663acf3b08cdecef6aa8a2c019ed647f02cb970f078d88e3ec7d",
+            "hendrycks_high_school_macroeconomics_PL_test.jsonl": "9425c91f59bd35355262f0c60b04b479a2e700064751efddadb9a28cbc83974d",
+            "hendrycks_high_school_mathematics_PL_test.jsonl": "80c73fa2b882104047260829a57cdca588d1ec9ba690b8a8a5cb87f65283e735",
+            "hendrycks_high_school_microeconomics_PL_test.jsonl": "aebc508914a9b32f5dd1bdc3ebd1375a619791214cb67968ccc6d5712675cbe8",
+            "hendrycks_high_school_physics_PL_test.jsonl": "bba7751511cada0b7b0b07133f2a961e5f046e174e201628a6722a279ab4e47f",
+            "hendrycks_high_school_psychology_PL_test.jsonl": "adf5943b42493ae59f9310028a45adffaace2359d52b8cbb1fa9cc41d6371e45",
+            "hendrycks_high_school_statistics_PL_test.jsonl": "14ffdec0e35d7437d7b2243c3bcbf606843df9ac4eaa044cae95faa3f6e41fdf",
+            "hendrycks_high_school_us_history_PL_test.jsonl": "d8a7946dbf1860b1e99ff0d1948ede18500cc06984ef68786ffed6a8160731f7",
+            "hendrycks_high_school_world_history_PL_test.jsonl": "ae943acd33cf4b23fd3d70d4cc5b0d16f073574c547de91bfc157d552bf63ec0",
+            "hendrycks_human_aging_PL_test.jsonl": "21df1fd8d194a6e453e71e1583626fc723f1c9b44cc571ef7d6074450754ed82",
+            "hendrycks_human_sexuality_PL_test.jsonl": "d7c467fcee3671069986948b69e6ec69c2630173207025ce4925719bb1a880d4",
+            "hendrycks_international_law_PL_test.jsonl": "f0b6f86e54390b63fda0ecf07c7c55f87d30864f6a698c9819f8d8e6eca4d746",
+            "hendrycks_jurisprudence_PL_test.jsonl": "9d7f42f41510da5d835d8df834c4bf6bcd2e03e658cc906ec334810b30ebbe96",
+            "hendrycks_logical_fallacies_PL_test.jsonl": "8b7cd34e9126adcfbc4b9b7bee8345d97352ec892221a0e98fa268210b04ee51",
+            "hendrycks_machine_learning_PL_test.jsonl": "f654c60a51877a9d269471ab01b54b0d4d20f9cc07f7fb57512fb3e90d5a8608",
+            "hendrycks_management_PL_test.jsonl": "628374b8f9beabefe43c2b1dda2c07760eb455d2e15f9e58f58506d61f0fc3bb",
+            "hendrycks_marketing_PL_test.jsonl": "31dc1eef1527c98642080c0b3866ea113bda77c2a03c5155aafe382206ffacdb",
+            "hendrycks_medical_genetics_PL_test.jsonl": "f05fcc33d029460a7f7916ec83debfc746ced5ea931f7a56b63fe1245819e4a4",
+            "hendrycks_miscellaneous_PL_test.jsonl": "94be7854c05aef25ebca4ae1dc562f37172b5346a14fb760b8e89eefea588809",
+            "hendrycks_moral_disputes_PL_test.jsonl": "245ad1cf635274e5ac81559542cb21638b7c729e8a811f76271f4301937d6127",
+            "hendrycks_moral_scenarios_PL_test.jsonl": "d60b08bb71451594385db5ff6ffdef4ca11fec801dc86d681e68f1727a690f9b",
+            "hendrycks_nutrition_PL_test.jsonl": "3cfe618d8e23e3c8e5c752544f07a71b90b4c84d4bc5d673026e8be4de4bf285",
+            "hendrycks_philosophy_PL_test.jsonl": "a1939f9773085bdae7e127fce099afc4153ec4cbe5f66e126f12599feaf76639",
+            "hendrycks_prehistory_PL_test.jsonl": "42eaf871b859f22321dfdec21b6b6b4a3c8f1435c2d4cd2da46dc9fb5ceced88",
+            "hendrycks_professional_accounting_PL_test.jsonl": "1fb6440616a195b94d3de760af5cacd05656667dfac595e2601504cda407c25e",
+            "hendrycks_professional_law_PL_test.jsonl": "fb86b175ecdfe8b56d4991bf6d6705a58f1d89a3efda2a8eb5f01f8082b21e57",
+            "hendrycks_professional_medicine_PL_test.jsonl": "ec818eb13a66e631ab82a10bae1536377dcd4bcadf2321c964b7718c60992080",
+            "hendrycks_professional_psychology_PL_test.jsonl": "72ddab1aaddc26266f4964941334a84d9243f63c329d2d9349bdcaa20efa6bc3",
+            "hendrycks_public_relations_PL_test.jsonl": "2e16835d88ba1ebb6f70a526bd85b9830cb1d3e14e2358a43ce15a3d548232e1",
+            "hendrycks_security_studies_PL_test.jsonl": "c476a634f4f48ff025869158057de726a8080bee253d88921d90252204aa3e4c",
+            "hendrycks_sociology_PL_test.jsonl": "dd56860da95ed793943c9c201f9aa8f5aa8d83753b1bbcbc440fff7ac30f881a",
+            "hendrycks_us_foreign_policy_PL_test.jsonl": "95183b3fcd02d17f5bb7d359b709b3fde13bf6f0f0f1375ab676b0e7dda36786",
+            "hendrycks_virology_PL_test.jsonl": "99e2c6f80907315317619f8c92f8d501ae484619c1169e36d9ad8fa59abfbda7",
+            "hendrycks_world_religions_PL_test.jsonl": "4631dbc58d379cbc46751515f36af682f8bf53db36f3345dbe090d5a00b7b24a"
+          },
+          "acc": 0.235009,
+          "acc_ci95": [
+            0.227959,
+            0.24206
+          ],
+          "acc_norm": 0.235009,
+          "acc_norm_ci95": [
+            0.22803,
+            0.242273
+          ],
+          "split": "test; micro-average over subjects",
+          "correct_primary": 3300
+        }
+      }
+    },
+    "1.09-k20000": {
+      "step": 20000,
+      "weights_sha256": "8bb1927755ae856a5f9f9e63ac2d3f94cd750a8059eeca81be86cf820abc2461",
+      "weights_published": false,
+      "source_results_sha256": "a6225b2af295c846852314f88c0e08f3a71d67d3cd67a53c65bdd87b989a018f",
+      "evaluator_sha256": "ae6658f9c83a484e373dc6d14476a7aeb98a905c27cd6d97ea234fb56140e016",
+      "tasks": {
+        "arc_challenge": {
+          "n": 1172,
+          "n_pelne": 1172,
+          "dataset_repo": "allenai/ai2_arc",
+          "dataset_sha": "210d026faf9955653af8916fad021475a3f00453",
+          "pliki_sha256": {
+            "ARC-Challenge/test-00000-of-00001.parquet": "62f03257e737aed263f55c6abf87c7bb0028a44a6bdd2a26eb1279eb42c1d1e9"
+          },
+          "acc": 0.306314,
+          "acc_ci95": [
+            0.280717,
+            0.332765
+          ],
+          "acc_norm": 0.331058,
+          "acc_norm_ci95": [
+            0.304608,
+            0.358362
+          ],
+          "split": "test",
+          "correct_primary": 388
+        },
+        "hellaswag": {
+          "n": 10042,
+          "n_pelne": 10042,
+          "dataset_repo": "Rowan/hellaswag",
+          "dataset_sha": "218ec52e09a7e7462a5400043bb9a69a41d06b76",
+          "pliki_sha256": {
+            "data/validation-00000-of-00001.parquet": "899813071e1e95efafec90f856e1987d2150fa4d020fc005df6962c259f660cd"
+          },
+          "acc": 0.397232,
+          "acc_ci95": [
+            0.387473,
+            0.406592
+          ],
+          "acc_norm": 0.503784,
+          "acc_norm_ci95": [
+            0.493926,
+            0.513543
+          ],
+          "split": "validation",
+          "correct_primary": 5059
+        },
+        "piqa": {
+          "n": 1838,
+          "n_pelne": 1838,
+          "dataset_repo": "baber/piqa",
+          "dataset_sha": "142f6d7367fd9877f0fb3b5734ea6a545f54cdd1",
+          "pliki_sha256": {
+            "piqa_validation.parquet": "2b17ee1215b1ceb2ddabf4c90a05c5a8b833bef9d207960f57d111c43a2080f3"
+          },
+          "acc": 0.702938,
+          "acc_ci95": [
+            0.682263,
+            0.723069
+          ],
+          "acc_norm": 0.707835,
+          "acc_norm_ci95": [
+            0.68716,
+            0.728509
+          ],
+          "split": "validation",
+          "correct_primary": 1301
+        },
+        "winogrande": {
+          "n": 1267,
+          "n_pelne": 1267,
+          "dataset_repo": "allenai/winogrande",
+          "dataset_sha": "01e74176c63542e6b0bcb004dcdea22d94fb67b5",
+          "pliki_sha256": {
+            "winogrande_xl/validation-00000-of-00001.parquet": "f9d914d1818c0aba98cabdf8af2bc4bd943c462b9c5e7062647e05c2ce3315d1"
+          },
+          "acc": 0.550118,
+          "acc_ci95": [
+            0.522494,
+            0.577743
+          ],
+          "acc_norm": 0.550118,
+          "acc_norm_ci95": [
+            0.523283,
+            0.577743
+          ],
+          "split": "validation",
+          "correct_primary": 697
+        },
+        "boolq": {
+          "n": 3270,
+          "n_pelne": 3270,
+          "dataset_repo": "aps/super_glue",
+          "dataset_sha": "3de24cf8022e94f4ee4b9d55a6f539891524d646",
+          "pliki_sha256": {
+            "boolq/validation-00000-of-00001.parquet": "f46438b8cce048184552a2bfcdb8252eadb3a2602f0c22334caccd9b6ea547d6"
+          },
+          "acc": 0.514067,
+          "acc_ci95": [
+            0.496942,
+            0.531193
+          ],
+          "acc_norm": 0.617737,
+          "acc_norm_ci95": [
+            0.601223,
+            0.634564
+          ],
+          "split": "validation",
+          "correct_primary": 1681
+        },
+        "openbookqa": {
+          "n": 500,
+          "n_pelne": 500,
+          "dataset_repo": "allenai/openbookqa",
+          "dataset_sha": "388097ea7776314e93a529163e0fea805b8a6454",
+          "pliki_sha256": {
+            "main/test-00000-of-00001.parquet": "cd5483e366daa230c1c87bbdc512d8b7229f14f6dd04d19fc8b1a3855aaaa8a3"
+          },
+          "acc": 0.244,
+          "acc_ci95": [
+            0.208,
+            0.282
+          ],
+          "acc_norm": 0.352,
+          "acc_norm_ci95": [
+            0.31,
+            0.394
+          ],
+          "split": "test",
+          "correct_primary": 176
+        },
+        "belebele_pol": {
+          "n": 900,
+          "n_pelne": 900,
+          "dataset_repo": "facebook/belebele",
+          "dataset_sha": "7899cdfa4e1e0d733fd77c848e2c273cb1d32be2",
+          "pliki_sha256": {
+            "data/pol_Latn.jsonl": "1699bb60b65e258d71989c541027c5439eb36a4f5032a5be392137bf5a1b6bec"
+          },
+          "acc": 0.246667,
+          "acc_ci95": [
+            0.218889,
+            0.275556
+          ],
+          "acc_norm": 0.246667,
+          "acc_norm_ci95": [
+            0.218889,
+            0.275556
+          ],
+          "split": "pol_Latn evaluation set",
+          "correct_primary": 222
+        },
+        "hellaswag_pl": {
+          "n": 10042,
+          "n_pelne": 10042,
+          "dataset_repo": "Eurolingua/hellaswagx",
+          "dataset_sha": "c38911a204759866dbcde35eeb584c36caf11a8b",
+          "pliki_sha256": {
+            "hellaswag_PL_validation.jsonl": "acdfc6d397885a4c86de178a6e5f9b7e9b90a7ff81ffcd6e5e0d3d0788dd3a16"
+          },
+          "acc": 0.345549,
+          "acc_ci95": [
+            0.336288,
+            0.355009
+          ],
+          "acc_norm": 0.442442,
+          "acc_norm_ci95": [
+            0.432583,
+            0.452201
+          ],
+          "split": "validation",
+          "correct_primary": 4443
+        },
+        "mmlu_pl": {
+          "n": 14042,
+          "n_pelne": 14042,
+          "dataset_repo": "Eurolingua/mmlux",
+          "dataset_sha": "7fb62cf05c1ce9ea550dbfd2ea6c145204344882",
+          "pliki_sha256": {
+            "hendrycks_abstract_algebra_PL_test.jsonl": "4905c6ef8d306c2ff1e1771ef8c15b63152bb88effec11844217893b81fcc3ea",
+            "hendrycks_anatomy_PL_test.jsonl": "5cba9bc54a6f0bdaddcd2e0e23155fccfeab469b561f03cc571e1387abeb9a38",
+            "hendrycks_astronomy_PL_test.jsonl": "00dc5fca7a8be3d8bef9e24ad9018e5417fb07703ebd2d0039c2c82418272b03",
+            "hendrycks_business_ethics_PL_test.jsonl": "9c9542880c720e2d0f5a768ab2d6602bdf5c394d66e51ff352af79aaca6bbd96",
+            "hendrycks_clinical_knowledge_PL_test.jsonl": "79fd5fd7a17dc7c7d84a9e9b138cf13c2dcaf6ab332eb1a0f299a7b52dc89c61",
+            "hendrycks_college_biology_PL_test.jsonl": "b49d766fb2c4160809d3907b846cd6c5c609299f47d9096bcf3208b843392c48",
+            "hendrycks_college_chemistry_PL_test.jsonl": "36c6fd391411789310dca0bbd36aec9aae8784df2bac2e6838e0895eebd97f1d",
+            "hendrycks_college_computer_science_PL_test.jsonl": "90ab29151dc4b8c7e13893516474201b73e8f9765cdf87225b6b9cacdd540e70",
+            "hendrycks_college_mathematics_PL_test.jsonl": "84411ee6464d01f24f2482bccb451635fe75d8169e33932ce6e067c1be3d37ee",
+            "hendrycks_college_medicine_PL_test.jsonl": "faa451d891c8a67b15bae4921c49a8afea448ac7b1b6095646c3570ac3040606",
+            "hendrycks_college_physics_PL_test.jsonl": "bfcaac4cb6d13e7cbcc6ebfb7ea6a9ab0d0ec07613b1b10ca679e632106b4bc2",
+            "hendrycks_computer_security_PL_test.jsonl": "c8324fb94b5b4b7e17440520ea20a22d8c03c3b86da81824059602982463cf8a",
+            "hendrycks_conceptual_physics_PL_test.jsonl": "5227e303a5437f437a4ad9d7eb75917d36051ce635af9f43a92d9525918eac64",
+            "hendrycks_econometrics_PL_test.jsonl": "9bf1d2a2c9eb4f5692b3876919dbc4e679d43d3ea2b64dc8fa2c17703a3e696b",
+            "hendrycks_electrical_engineering_PL_test.jsonl": "a431be1d342cc1eca6aa408688b78751c74621056a4fa4c31e7dfda443e4be6f",
+            "hendrycks_elementary_mathematics_PL_test.jsonl": "9cdb8692fc08d6d8ac5dec24d89f386863b6a02480eb3b8c5be641f15700e667",
+            "hendrycks_formal_logic_PL_test.jsonl": "0f30d48346a0cd2f136463b283d8bd3fe085de67da2a34d90cac689740f9b5be",
+            "hendrycks_global_facts_PL_test.jsonl": "21c753017550f0d31d7c2cdbe82637c877682a9edf8ab913de720e79c82b374f",
+            "hendrycks_high_school_biology_PL_test.jsonl": "7b086933dd97cfc392fba5f78c3523ddc0302cf2b7caed7428a75bfdf9fbb7f2",
+            "hendrycks_high_school_chemistry_PL_test.jsonl": "85f836040875313d60cf9241bd9e827a3cdb770cf159ddf3f7032edb293871ec",
+            "hendrycks_high_school_computer_science_PL_test.jsonl": "034b7b497ada5cc1146c6c5fa02291094d5da512bc0c618ae2b967bd941a70c1",
+            "hendrycks_high_school_european_history_PL_test.jsonl": "023ce8f773cb5dccd86658165a86a5bc1b0d83d53cede5a22c8f7a027582e2a4",
+            "hendrycks_high_school_geography_PL_test.jsonl": "3cce3d6defc5cad9742f9db4407ec5176e6b5d1634b0f8ecb4cfe9f4cabfbe5a",
+            "hendrycks_high_school_government_and_politics_PL_test.jsonl": "9c73c16602cd663acf3b08cdecef6aa8a2c019ed647f02cb970f078d88e3ec7d",
+            "hendrycks_high_school_macroeconomics_PL_test.jsonl": "9425c91f59bd35355262f0c60b04b479a2e700064751efddadb9a28cbc83974d",
+            "hendrycks_high_school_mathematics_PL_test.jsonl": "80c73fa2b882104047260829a57cdca588d1ec9ba690b8a8a5cb87f65283e735",
+            "hendrycks_high_school_microeconomics_PL_test.jsonl": "aebc508914a9b32f5dd1bdc3ebd1375a619791214cb67968ccc6d5712675cbe8",
+            "hendrycks_high_school_physics_PL_test.jsonl": "bba7751511cada0b7b0b07133f2a961e5f046e174e201628a6722a279ab4e47f",
+            "hendrycks_high_school_psychology_PL_test.jsonl": "adf5943b42493ae59f9310028a45adffaace2359d52b8cbb1fa9cc41d6371e45",
+            "hendrycks_high_school_statistics_PL_test.jsonl": "14ffdec0e35d7437d7b2243c3bcbf606843df9ac4eaa044cae95faa3f6e41fdf",
+            "hendrycks_high_school_us_history_PL_test.jsonl": "d8a7946dbf1860b1e99ff0d1948ede18500cc06984ef68786ffed6a8160731f7",
+            "hendrycks_high_school_world_history_PL_test.jsonl": "ae943acd33cf4b23fd3d70d4cc5b0d16f073574c547de91bfc157d552bf63ec0",
+            "hendrycks_human_aging_PL_test.jsonl": "21df1fd8d194a6e453e71e1583626fc723f1c9b44cc571ef7d6074450754ed82",
+            "hendrycks_human_sexuality_PL_test.jsonl": "d7c467fcee3671069986948b69e6ec69c2630173207025ce4925719bb1a880d4",
+            "hendrycks_international_law_PL_test.jsonl": "f0b6f86e54390b63fda0ecf07c7c55f87d30864f6a698c9819f8d8e6eca4d746",
+            "hendrycks_jurisprudence_PL_test.jsonl": "9d7f42f41510da5d835d8df834c4bf6bcd2e03e658cc906ec334810b30ebbe96",
+            "hendrycks_logical_fallacies_PL_test.jsonl": "8b7cd34e9126adcfbc4b9b7bee8345d97352ec892221a0e98fa268210b04ee51",
+            "hendrycks_machine_learning_PL_test.jsonl": "f654c60a51877a9d269471ab01b54b0d4d20f9cc07f7fb57512fb3e90d5a8608",
+            "hendrycks_management_PL_test.jsonl": "628374b8f9beabefe43c2b1dda2c07760eb455d2e15f9e58f58506d61f0fc3bb",
+            "hendrycks_marketing_PL_test.jsonl": "31dc1eef1527c98642080c0b3866ea113bda77c2a03c5155aafe382206ffacdb",
+            "hendrycks_medical_genetics_PL_test.jsonl": "f05fcc33d029460a7f7916ec83debfc746ced5ea931f7a56b63fe1245819e4a4",
+            "hendrycks_miscellaneous_PL_test.jsonl": "94be7854c05aef25ebca4ae1dc562f37172b5346a14fb760b8e89eefea588809",
+            "hendrycks_moral_disputes_PL_test.jsonl": "245ad1cf635274e5ac81559542cb21638b7c729e8a811f76271f4301937d6127",
+            "hendrycks_moral_scenarios_PL_test.jsonl": "d60b08bb71451594385db5ff6ffdef4ca11fec801dc86d681e68f1727a690f9b",
+            "hendrycks_nutrition_PL_test.jsonl": "3cfe618d8e23e3c8e5c752544f07a71b90b4c84d4bc5d673026e8be4de4bf285",
+            "hendrycks_philosophy_PL_test.jsonl": "a1939f9773085bdae7e127fce099afc4153ec4cbe5f66e126f12599feaf76639",
+            "hendrycks_prehistory_PL_test.jsonl": "42eaf871b859f22321dfdec21b6b6b4a3c8f1435c2d4cd2da46dc9fb5ceced88",
+            "hendrycks_professional_accounting_PL_test.jsonl": "1fb6440616a195b94d3de760af5cacd05656667dfac595e2601504cda407c25e",
+            "hendrycks_professional_law_PL_test.jsonl": "fb86b175ecdfe8b56d4991bf6d6705a58f1d89a3efda2a8eb5f01f8082b21e57",
+            "hendrycks_professional_medicine_PL_test.jsonl": "ec818eb13a66e631ab82a10bae1536377dcd4bcadf2321c964b7718c60992080",
+            "hendrycks_professional_psychology_PL_test.jsonl": "72ddab1aaddc26266f4964941334a84d9243f63c329d2d9349bdcaa20efa6bc3",
+            "hendrycks_public_relations_PL_test.jsonl": "2e16835d88ba1ebb6f70a526bd85b9830cb1d3e14e2358a43ce15a3d548232e1",
+            "hendrycks_security_studies_PL_test.jsonl": "c476a634f4f48ff025869158057de726a8080bee253d88921d90252204aa3e4c",
+            "hendrycks_sociology_PL_test.jsonl": "dd56860da95ed793943c9c201f9aa8f5aa8d83753b1bbcbc440fff7ac30f881a",
+            "hendrycks_us_foreign_policy_PL_test.jsonl": "95183b3fcd02d17f5bb7d359b709b3fde13bf6f0f0f1375ab676b0e7dda36786",
+            "hendrycks_virology_PL_test.jsonl": "99e2c6f80907315317619f8c92f8d501ae484619c1169e36d9ad8fa59abfbda7",
+            "hendrycks_world_religions_PL_test.jsonl": "4631dbc58d379cbc46751515f36af682f8bf53db36f3345dbe090d5a00b7b24a"
+          },
+          "acc": 0.238641,
+          "acc_ci95": [
+            0.231662,
+            0.245763
+          ],
+          "acc_norm": 0.238641,
+          "acc_norm_ci95": [
+            0.231591,
+            0.245834
+          ],
+          "split": "test; micro-average over subjects",
+          "correct_primary": 3351
+        }
+      }
+    }
+  },
+  "additional_measurements": {
+    "benchteam_gsm8k_sst2_20261111": {
+      "measurement_owner": "SlayerLab / Fabryka AI",
+      "status": "descriptive; training-lineage contamination scan pending",
+      "measured_utc_date": "2026-11-11",
+      "source_manifest_sha256": "229c86053bda604716282e314e584f2eac24d870dc5f99e102328536ecd9d95e",
+      "protocol": {
+        "source_repository": "https://github.com/slayerlabs/benchmarks-fabryka-ai",
+        "inventory_commit": "8b547007526971bd2fc35dfe56ff2c7d9b802568",
+        "source_protocol_id": "bench50-v0.3.1-fullB",
+        "scope": "two-task subset, full evaluated splits; not a full 50-task result",
+        "model_adapter": "slayer",
+        "lm_eval_version": "0.4.9.1",
+        "dtype": "float32",
+        "batch_size": 1,
+        "context_limit": 4096,
+        "task_settings": {
+          "gsm8k": {
+            "task_id": "T33",
+            "split": "test",
+            "few_shot": 5,
+            "metric": "exact_match,strict-match"
+          },
+          "sst2": {
+            "task_id": "T21",
+            "split": "validation",
+            "few_shot": 0,
+            "metric": "acc,none"
+          }
+        },
+        "metric_units": "fraction; multiply by 100 for card percentages",
+        "ci_interpretation": "pointwise intervals for individual scores as reported; not confidence intervals for model differences or training-seed variance",
+        "dataset_identity": "source_id_digest pins the evaluated item set; dataset revisions are not supplied in these aggregate reports"
+      },
+      "models": {
+        "1.07": {
+          "step": 22327,
+          "weights_published": true,
+          "repo": "SlayerLab/Slayer1B-1.07",
+          "revision": "a090b54e658fb57dc5166c1fcc90cc6bf5c80106",
+          "weights_sha256": "d826b4a11bf4eaccee10fe43ec19a4f6d515dfc5f3e5c859adb631f1a0b76ad5",
+          "source_run_id": "niezalezny-SlayerLab_Slayer1B-1.07-a090b54e-20261111T012858Z",
+          "source_aggregate_sha256": "4639f14fc11b2c254f91fde7b7f606cd90a8480a443df37c6671efb3c09c8c7b",
+          "tasks": {
+            "gsm8k": {
+              "task_id": "T33",
+              "metric": "exact_match,strict-match",
+              "value": 0.015163002274450341,
+              "unit": "score",
+              "ci95": {
+                "lower": 0.009097801364670205,
+                "upper": 0.021986353297952996,
+                "level": 0.95,
+                "method": "paired_bootstrap",
+                "replicates": 10000,
+                "seed": 20261028
+              },
+              "n_scored": 1319,
+              "n_population": 1319,
+              "contamination": "unscanned",
+              "state": "ok",
+              "runner_sha256": "c906036b4e597e2c17ad7f1f6441e72fbf8e066946c2a7731966612674e3995e",
+              "results_sha256": "d20e0d2b1f0e21915f857c6c4218f43f4c1db4c2799e28a78c3618d4fcfcf1b6",
+              "source_id_digest": "a32d2c4b15b4e22c87fc4e4717cf476baa2aa90a17f61230cef747c0a3bb61d0"
+            },
+            "sst2": {
+              "task_id": "T21",
+              "metric": "acc,none",
+              "value": 0.5103211009174312,
+              "unit": "score",
+              "ci95": {
+                "lower": 0.47706422018348627,
+                "upper": 0.5435779816513762,
+                "level": 0.95,
+                "method": "paired_bootstrap",
+                "replicates": 10000,
+                "seed": 20261028
+              },
+              "n_scored": 872,
+              "n_population": 872,
+              "contamination": "unscanned",
+              "state": "ok",
+              "runner_sha256": "c906036b4e597e2c17ad7f1f6441e72fbf8e066946c2a7731966612674e3995e",
+              "results_sha256": "f1965fb1a3651c6ab526a82d9bc35504b8c4f92c4789b471d61fa6a37e54c83f",
+              "source_id_digest": "7a716d60452e40d779aaad8da40aa11cdfb66cf0a069b6f0a5c5b38c78819ce7"
+            }
+          }
+        },
+        "1.08": {
+          "step": 1709,
+          "weights_published": true,
+          "repo": "SlayerLab/Slayer1B-1.08",
+          "revision": "5f854a0cc7160594d54d1788d37a44e8ebabe16c",
+          "weights_sha256": "1c510d9a59d3b9e62360d1db57e08fa450fb9b09df849df7f2eef132ea6715a4",
+          "source_run_id": "niezalezny-SlayerLab_Slayer1B-1.08-5f854a0c-20261111T023631Z",
+          "source_aggregate_sha256": "3ee230fb5029cf9c6424c27bae2b60dfd345998b2bee42021cb53e625021615b",
+          "tasks": {
+            "gsm8k": {
+              "task_id": "T33",
+              "metric": "exact_match,strict-match",
+              "value": 0.009097801364670205,
+              "unit": "score",
+              "ci95": {
+                "lower": 0.004548900682335102,
+                "upper": 0.014404852160727824,
+                "level": 0.95,
+                "method": "paired_bootstrap",
+                "replicates": 10000,
+                "seed": 20261028
+              },
+              "n_scored": 1319,
+              "n_population": 1319,
+              "contamination": "unscanned",
+              "state": "ok",
+              "runner_sha256": "c906036b4e597e2c17ad7f1f6441e72fbf8e066946c2a7731966612674e3995e",
+              "results_sha256": "7e2f64e0958f9e5897debe2804cfb25c804b4abe1b8bdd440f7885d16b98f3e6",
+              "source_id_digest": "a32d2c4b15b4e22c87fc4e4717cf476baa2aa90a17f61230cef747c0a3bb61d0"
+            },
+            "sst2": {
+              "task_id": "T21",
+              "metric": "acc,none",
+              "value": 0.5045871559633027,
+              "unit": "score",
+              "ci95": {
+                "lower": 0.4713302752293578,
+                "upper": 0.536697247706422,
+                "level": 0.95,
+                "method": "paired_bootstrap",
+                "replicates": 10000,
+                "seed": 20261028
+              },
+              "n_scored": 872,
+              "n_population": 872,
+              "contamination": "unscanned",
+              "state": "ok",
+              "runner_sha256": "c906036b4e597e2c17ad7f1f6441e72fbf8e066946c2a7731966612674e3995e",
+              "results_sha256": "1e1a7b5d6c41dff7140208736213f308a5be62986bef1e9bf5b7aeb3e176a48a",
+              "source_id_digest": "7a716d60452e40d779aaad8da40aa11cdfb66cf0a069b6f0a5c5b38c78819ce7"
+            }
+          }
+        },
+        "1.09-k10000": {
+          "step": 10000,
+          "weights_published": false,
+          "repo": null,
+          "revision": null,
+          "weights_sha256": "12939d84ef0d06b350773fb30837390306d899ded315d1721b52c7615a87317a",
+          "source_run_id": "niezalezny-Slayer1B-1.09-k10000-lokalny-lokalny-20261111T025334Z",
+          "source_aggregate_sha256": "042b2639bc325aeb7fb60787bc8205d1384557e37dbb7b01ba5489db41ebdfe8",
+          "tasks": {
+            "gsm8k": {
+              "task_id": "T33",
+              "metric": "exact_match,strict-match",
+              "value": 0.009855951478392721,
+              "unit": "score",
+              "ci95": {
+                "lower": 0.004548900682335102,
+                "upper": 0.015163002274450341,
+                "level": 0.95,
+                "method": "paired_bootstrap",
+                "replicates": 10000,
+                "seed": 20261028
+              },
+              "n_scored": 1319,
+              "n_population": 1319,
+              "contamination": "unscanned",
+              "state": "ok",
+              "runner_sha256": "c906036b4e597e2c17ad7f1f6441e72fbf8e066946c2a7731966612674e3995e",
+              "results_sha256": "5cf7805fa9c7d3d65d6af44920b6659287c621294efe4d7cde62f2055e65562e",
+              "source_id_digest": "a32d2c4b15b4e22c87fc4e4717cf476baa2aa90a17f61230cef747c0a3bb61d0"
+            },
+            "sst2": {
+              "task_id": "T21",
+              "metric": "acc,none",
+              "value": 0.5825688073394495,
+              "unit": "score",
+              "ci95": {
+                "lower": 0.5493119266055045,
+                "upper": 0.6135321100917431,
+                "level": 0.95,
+                "method": "paired_bootstrap",
+                "replicates": 10000,
+                "seed": 20261028
+              },
+              "n_scored": 872,
+              "n_population": 872,
+              "contamination": "unscanned",
+              "state": "ok",
+              "runner_sha256": "c906036b4e597e2c17ad7f1f6441e72fbf8e066946c2a7731966612674e3995e",
+              "results_sha256": "d47706c62088fe772f1c26d2ed13f52dfecc8d832bbdecd14dd24c89086e9659",
+              "source_id_digest": "7a716d60452e40d779aaad8da40aa11cdfb66cf0a069b6f0a5c5b38c78819ce7"
+            }
+          }
+        }
+      }
+    }
+  },
+  "source_generator_note": "The updated approved aggregate does not supply its CI-generator SHA; evaluator SHA is retained per model. Previous aggregate: ce0822a324f805bb61f28cc8ea3197e6a9abd97a4591239349dc60141f5289db; previous CI generator: b792b5a7b4f5218acff4b3d5e9fb63c6f11d39fa26cca6fbd467758c90555a34."
 }

README.md CHANGED Viewed

@@ -16,14 +16,15 @@ tags:
 **Arkadiusz Słota, Anna Sieczka, Piotr Styła / Fabryka AI (SlayerLab)**
-**Execution snapshot, 10 November 2026, 21:35 UTC: continued pretraining from Slayer1B 1.08 has resumed on eight H200 GPUs. The objective remains useful training exposure within a total USD 425 campaign budget.** After the earlier cost pause, operators verified fresh numerical control, preflight and `RESUME @ 11600` at 21:28 UTC; the update-11,610 loss matched the pre-stop value. The public trace subsequently reached update 11,690. The tracking plan is 40,000 updates, while the trainer still retains its 80,000-update scheduling argument during the constant-rate phase. Neither value is a promise of completed exposure; final length depends on the remaining budget and completion of learning-rate decay. This is not a completed model or weight release.
 | Verified execution evidence | Scope |
 |---|---|
 | CPT start | Approximately 04:50 UTC, 10 November 2026, on eight B200 GPUs |
-| Recovery on H200 | Operators verified fresh numerical control and `RESUME @ 9600`; training subsequently passed update 10,000 |
-| Local weights-only checkpoints | Updates 5,000 and 10,000, with transfer checksums verified |
-| Latest local full checkpoint | Update 10,000; includes optimizer state, retained separately from the weights-only checkpoints |
 The [CPT Track history](https://track.fabryka.ai/run/71cf98ec-7c60-54a8-ac47-0ea7a23f573a) provides the changing public telemetry. This card is an execution and budget snapshot, not a completed 1.09 model, final quality verdict or inference-ready weight release.
@@ -47,7 +48,7 @@ The [training configuration](TRAINING_CONFIGURATION.md) distinguishes the still-
 | Configured scheduling argument | Currently 80,000 updates; this keeps the WSD constant-rate phase flexible and is not a commitment to complete 83.89B new positions |
 | New-source allocation | **20% PL / 80% EN target**; not a measured cumulative model-language ratio |
 | Optimizer / schedule | Muon peak 0.02 plus AdamW peak 0.0006; 2,000-update warmup; final 20% WSD decay to 0.00006 |
-| Checkpoints | Full state every 400 updates after the approved cadence change; weights-only checkpoints at each 5,000-update boundary; update-10,000 copies verified locally |
 | Code | Launcher v5 SHA256 `796c71016c1e167bfd3ca98b78518f91c3502a96eac233dd7186ef6639fcc656`; exact command and converter pins in the configuration document |
 The reviewed inherited count is **18,863,257,682 source/file-disjoint positions**, not a globally content-deduplicated history. The measured c01 duplicate correction is conservatively **114,627,128 positions**, the sum of two exclusion-list counts rather than a measured union. If a 40,000-update endpoint is selected and completed, then 41,943,040,000 new positions would give a conditional source-position total of **60,691,670,554** after that correction. Historical content-level near-duplicates remain unresolved. This projection is not evidence of completed exposure or a globally unique corpus.
@@ -90,7 +91,7 @@ The last received public sample before the pause is update **450**, training los
 Fresh numerical control and step-zero `INIT-FROM` preflight were verified for the CPT start described above. The earlier permission for the abandoned slot was not reused as that evidence. Any later physical-host restart still requires its own numerical control; successful start does not prove a checkpoint restore.
-Working checkpoints are copied to a trusted local workstation, not published as Hugging Face weights. Verified weights-only checkpoints include updates 5,000 and 10,000. The latest full update-10,000 checkpoint has SHA256 `01ccb2bed25c43b43356c8a813f9bf3f8b55c481bce85087e60862001f1e3dea`; the weights-only update-10,000 file has SHA256 `fcfd2d3044d504706519fb6e1bd8d6dea959b162dac62e0f5967bbd6807df643`. Weights-only checkpoints are retained, while the full local recovery checkpoint is replaced by the next verified copy. This retention policy does not eliminate exposure to a cloud-disk failure between transfers.
 ## Data and exposure
@@ -100,35 +101,37 @@ The PL/EN balance will follow the approved fresh-data inventory. The user explic
 New material requires licensing, PII filtering, contamination scans, grouped held-outs and deduplication against the base model's full data ancestry. A 1% document holdout is not necessarily 1% of tokens. Historical exposure in the starting weights cannot be undone by filtering new inputs; retrospective benchmark hits must be handled explicitly in evaluation denominators.
-## Standard benchmark measurements — update 5,000
-We thank **BenchTeam / Fabryka AI** for benchmark implementations, protocol work and evaluation proposals. Its independent [full benchmark report](https://github.com/slayerlabs/benchmarks-fabryka-ai/blob/8b547007526971bd2fc35dfe56ff2c7d9b802568/research/results/full-run-2026-11/README.md) is distinct from the **SlayerLab measurements below**, published in the [CPT Track history](https://track.fabryka.ai/run/71cf98ec-7c60-54a8-ac47-0ea7a23f573a). Update 5,000 is an intermediate checkpoint, not the eventual budget-limited endpoint or an inference-ready weight release.
-The same nine full splits were scored zero-shot in float32 for 1.07 update 22,327, 1.08 update 1,709 and 1.09 update 5,000. Scores are percentages with **pointwise 95% confidence intervals**; the named metric is accuracy (`acc`) or length-normalized accuracy (`acc_norm`). WinoGrande uses partial-sentence likelihood; MMLU-PL is micro-averaged. Polish HellaSwag and MMLU use the OpenGPTX task definitions.
-| Task | Metric | n | 1.07 | 1.08 | 1.09, update 5,000 |
-|---|---|---:|---:|---:|---:|
-| ARC-Challenge | acc_norm | 1,172 | 27.30 [24.74, 29.86] | 30.03 [27.39, 32.68] | 29.35 [26.79, 32.00] |
-| HellaSwag | acc_norm | 10,042 | 43.38 [42.43, 44.34] | 43.42 [42.43, 44.39] | 46.53 [45.55, 47.52] |
-| PIQA | acc_norm | 1,838 | 66.97 [64.80, 69.15] | 66.21 [64.15, 68.34] | 68.28 [66.21, 70.40] |
-| WinoGrande | acc | 1,267 | 53.59 [50.83, 56.35] | 53.67 [50.91, 56.43] | 54.93 [52.17, 57.70] |
-| BoolQ | acc | 3,270 | 58.29 [56.61, 59.97] | 59.97 [58.26, 61.59] | 61.80 [60.18, 63.49] |
-| OpenBookQA | acc_norm | 500 | 33.60 [29.40, 37.80] | 34.80 [30.60, 39.00] | 35.60 [31.40, 40.00] |
-| Belebele, Polish | acc | 900 | 26.00 [23.11, 28.89] | 24.00 [21.22, 26.89] | 24.11 [21.33, 26.89] |
-| HellaSwag-PL | acc_norm | 10,042 | 43.83 [42.84, 44.79] | 43.90 [42.90, 44.87] | 44.73 [43.76, 45.71] |
-| MMLU-PL | acc | 14,042 | 24.01 [23.30, 24.71] | 23.04 [22.34, 23.74] | 24.10 [23.39, 24.82] |
-Intervals use the percentile bootstrap with 10,000 replicates, seed base 20261110 and a model-specific stream. **Items are treated as independent; subject, template and source clusters are not modeled.** These are intervals for individual results, not paired differences or training-seed variance; overlap is not a test of a difference.
 The [machine-readable aggregate record](EVALUATION.json) supplies exact values, intervals, counts, dataset revisions and file digests, checkpoint and evaluator identities, and metric selection, without prompts, answers or per-item predictions. ARC-Challenge, OpenBookQA and MMLU-PL use test splits; HellaSwag, PIQA, WinoGrande, BoolQ and HellaSwag-PL use validation splits; Belebele uses the full Polish evaluation set.
 **Descriptive results; contamination scan of the training lineage against these tasks is pending.** Belebele and MMLU-PL are near their four-choice 25% random baseline. No overall average, ranking or release threshold is inferred. Numbers from different evaluators must not be silently substituted: prompting, BOS handling, normalization and software versions can change the result.
 There is no replay-only control for this continuation, and initialization, exposure and schedule differ between checkpoints. These measurements do not isolate a causal effect of the new datasets or quantify training-seed variance. Earlier internal diagnostics are separate historical records, not a criterion for naming or publishing this model.
 ## Additional GSM8K and SST-2 measurements — update 10,000
-SlayerLab measured this two-task subset on 11 November 2026 using the [pinned BenchTeam code](https://github.com/slayerlabs/benchmarks-fabryka-ai/tree/8b547007526971bd2fc35dfe56ff2c7d9b802568), lm-eval 0.4.9.1, the `slayer` adapter, float32, batch 1 and context limit 4,096. These are full evaluated splits, not a full 50-task run or a copy of the community's independent report. **Only this two-task section uses 1.09 update 10,000; the nine-task table above remains update 5,000.** Update 10,000 is an internal working checkpoint, not a published or inference-ready 1.09 weight release.
 | Task | Metric / prompting | n | 1.07 | 1.08 | 1.09, update 10,000 |
 |---|---|---:|---:|---:|---:|
@@ -141,6 +144,6 @@ Scores are percentages with reported pointwise 95% bootstrap intervals (10,000 d
 ## Cost, licensing and limitations
-The total campaign ceiling is **USD 425**, including GPU allocations, preparation, interruptions, storage and evaluation tracked by the resource meter. On 10 November at 18:57 UTC, the meter reported **USD 141.85 spent**, leaving approximately **USD 283**; this is counter attribution, not a settled invoice or a permanently current balance. Hardware selection is based on effective cost per processed position and verified availability. The final update count is reviewed at stage boundaries and must be selected before its WSD decay boundary. No higher spending limit or guaranteed completion date is implied.
 The selected 1.08 base declares Apache-2.0; source-text rights and future publication still require separate approval. Privacy filters are heuristic; generated text may be inaccurate, biased, offensive or personal. No reasoning quality, instruction-following ability or professional reliability has been demonstrated for the revised 1.09.

 **Arkadiusz Słota, Anna Sieczka, Piotr Styła / Fabryka AI (SlayerLab)**
+**Execution snapshot, 11 November 2026, 10:45 UTC: continued pretraining from Slayer1B 1.08 is running on eight B200 spot GPUs after an H200 segment and checkpoint recoveries. The campaign ceiling remains USD 425.** Operators reported update 27,140 at this snapshot; this is progress telemetry, not the endpoint of the run. The public tracking plan remains 40,000 updates, while the trainer retains its 80,000-update scheduling argument during the constant-rate phase. The next budget-dependent length review is at update 30,000; final length must allow completion of learning-rate decay. Neither configured value is a promise of completed exposure. This is not a completed model or inference-ready weight release.
 | Verified execution evidence | Scope |
 |---|---|
 | CPT start | Approximately 04:50 UTC, 10 November 2026, on eight B200 GPUs |
+| Historical H200 recovery | Operators verified fresh numerical control and checkpoint restores during the earlier H200 segment |
+| Current compute at the dated snapshot | Eight B200 spot GPUs; training continued beyond update 27,000 |
+| Checksum-verified weights-only copies on the workstation | Updates 5,000, 10,000, 15,000, 20,000 and 25,000; not public weight releases |
+| Latest verified full-state workstation copy | Update 25,000; includes optimizer state, retained separately from weights-only copies; not a claim about the newest recovery checkpoint on the cloud disk |
 The [CPT Track history](https://track.fabryka.ai/run/71cf98ec-7c60-54a8-ac47-0ea7a23f573a) provides the changing public telemetry. This card is an execution and budget snapshot, not a completed 1.09 model, final quality verdict or inference-ready weight release.
 | Configured scheduling argument | Currently 80,000 updates; this keeps the WSD constant-rate phase flexible and is not a commitment to complete 83.89B new positions |
 | New-source allocation | **20% PL / 80% EN target**; not a measured cumulative model-language ratio |
 | Optimizer / schedule | Muon peak 0.02 plus AdamW peak 0.0006; 2,000-update warmup; final 20% WSD decay to 0.00006 |
+| Checkpoints | Full state every 400 updates after the approved cadence change; weights-only checkpoints at each 5,000-update boundary; workstation copies verified through update 25,000 |
 | Code | Launcher v5 SHA256 `796c71016c1e167bfd3ca98b78518f91c3502a96eac233dd7186ef6639fcc656`; exact command and converter pins in the configuration document |
 The reviewed inherited count is **18,863,257,682 source/file-disjoint positions**, not a globally content-deduplicated history. The measured c01 duplicate correction is conservatively **114,627,128 positions**, the sum of two exclusion-list counts rather than a measured union. If a 40,000-update endpoint is selected and completed, then 41,943,040,000 new positions would give a conditional source-position total of **60,691,670,554** after that correction. Historical content-level near-duplicates remain unresolved. This projection is not evidence of completed exposure or a globally unique corpus.
 Fresh numerical control and step-zero `INIT-FROM` preflight were verified for the CPT start described above. The earlier permission for the abandoned slot was not reused as that evidence. Any later physical-host restart still requires its own numerical control; successful start does not prove a checkpoint restore.
+Working checkpoints are copied to a trusted workstation, not published as Hugging Face weights. Operators independently verified weights-only copies at updates 5,000, 10,000, 15,000, 20,000 and 25,000, and a separate full-state workstation copy at update 25,000. These transfer records are distinct from the newer recovery checkpoints retained on the running node and from the evaluated updates in the benchmark table. Weights-only copies are retained, while the full workstation recovery copy is replaced by the next verified transfer. This retention policy does not eliminate exposure to a cloud-disk failure between transfers.
 ## Data and exposure
 New material requires licensing, PII filtering, contamination scans, grouped held-outs and deduplication against the base model's full data ancestry. A 1% document holdout is not necessarily 1% of tokens. Historical exposure in the starting weights cannot be undone by filtering new inputs; retrospective benchmark hits must be handled explicitly in evaluation denominators.
+## Standard benchmark measurements — updates 5,000–20,000
+We thank **BenchTeam / Fabryka AI** for benchmark implementations, protocol work and evaluation proposals. Its independent [full benchmark report](https://github.com/slayerlabs/benchmarks-fabryka-ai/blob/8b547007526971bd2fc35dfe56ff2c7d9b802568/research/results/full-run-2026-11/README.md) is distinct from the **SlayerLab measurements below**, published in the [CPT Track history](https://track.fabryka.ai/run/71cf98ec-7c60-54a8-ac47-0ea7a23f573a). Updates 5,000, 10,000, 15,000 and 20,000 are distinct intermediate working checkpoints of the ongoing 1.09 run, not its eventual budget-limited endpoint or inference-ready weight releases.
+The same nine full splits were scored zero-shot in float32 for 1.07 update 22,327, 1.08 update 1,709 and 1.09 updates 5,000, 10,000, 15,000 and 20,000. Evaluator SHA, dataset revisions, file digests, task protocols and sample counts match across these rows. Scores are percentages with **pointwise 95% confidence intervals**; the named metric is accuracy (`acc`) or length-normalized accuracy (`acc_norm`). WinoGrande uses partial-sentence likelihood; MMLU-PL is micro-averaged. Polish HellaSwag and MMLU use the OpenGPTX task definitions.
+| Task | Metric | n | 1.07 | 1.08 | 1.09, update 5,000 | 1.09, update 10,000 | 1.09, update 15,000 | 1.09, update 20,000 |
+|---|---|---:|---:|---:|---:|---:|---:|---:|
+| ARC-Challenge | acc_norm | 1,172 | 27.30 [24.74, 29.86] | 30.03 [27.39, 32.68] | 29.35 [26.79, 32.00] | 30.46 [27.82, 33.11] | 32.51 [29.86, 35.24] | 33.11 [30.46, 35.84] |
+| HellaSwag | acc_norm | 10,042 | 43.38 [42.43, 44.34] | 43.42 [42.43, 44.39] | 46.53 [45.55, 47.52] | 48.93 [47.97, 49.91] | 49.77 [48.80, 50.76] | 50.38 [49.39, 51.35] |
+| PIQA | acc_norm | 1,838 | 66.97 [64.80, 69.15] | 66.21 [64.15, 68.34] | 68.28 [66.21, 70.40] | 68.99 [66.92, 71.11] | 68.99 [66.87, 71.06] | 70.78 [68.72, 72.85] |
+| WinoGrande | acc | 1,267 | 53.59 [50.83, 56.35] | 53.67 [50.91, 56.43] | 54.93 [52.17, 57.70] | 54.93 [52.17, 57.62] | 56.75 [54.06, 59.43] | 55.01 [52.25, 57.77] |
+| BoolQ | acc | 3,270 | 58.29 [56.61, 59.97] | 59.97 [58.26, 61.59] | 61.80 [60.18, 63.49] | 61.13 [59.45, 62.78] | 59.88 [58.20, 61.53] | 51.41 [49.69, 53.12] |
+| OpenBookQA | acc_norm | 500 | 33.60 [29.40, 37.80] | 34.80 [30.60, 39.00] | 35.60 [31.40, 40.00] | 34.40 [30.20, 38.60] | 36.00 [31.80, 40.20] | 35.20 [31.00, 39.40] |
+| Belebele, Polish | acc | 900 | 26.00 [23.11, 28.89] | 24.00 [21.22, 26.89] | 24.11 [21.33, 26.89] | 24.11 [21.33, 27.00] | 22.56 [19.89, 25.33] | 24.67 [21.89, 27.56] |
+| HellaSwag-PL | acc_norm | 10,042 | 43.83 [42.84, 44.79] | 43.90 [42.90, 44.87] | 44.73 [43.76, 45.71] | 45.51 [44.51, 46.48] | 44.18 [43.21, 45.16] | 44.24 [43.26, 45.22] |
+| MMLU-PL | acc | 14,042 | 24.01 [23.30, 24.71] | 23.04 [22.34, 23.74] | 24.10 [23.39, 24.82] | 23.32 [22.62, 24.03] | 23.50 [22.80, 24.21] | 23.86 [23.17, 24.58] |
+Intervals use the percentile bootstrap with 10,000 replicates and seed 20261110. The updated approved aggregate does not supply the random-stream derivation or CI-generator SHA; the previous aggregate documented model-specific streams. **Items are treated as independent; subject, template and source clusters are not modeled.** These are intervals for individual results, not paired differences or training-seed variance; overlap is not a test of a difference.
 The [machine-readable aggregate record](EVALUATION.json) supplies exact values, intervals, counts, dataset revisions and file digests, checkpoint and evaluator identities, and metric selection, without prompts, answers or per-item predictions. ARC-Challenge, OpenBookQA and MMLU-PL use test splits; HellaSwag, PIQA, WinoGrande, BoolQ and HellaSwag-PL use validation splits; Belebele uses the full Polish evaluation set.
+The approved aggregate SHA-256 is `fc6f67a20bdc6ada687349d35de26f6811e27766dfe1aa197082000c03535796`. Its earlier 1.07, 1.08 and 1.09 update-5,000 scores, intervals and dataset pins match the preceding aggregate; the three later checkpoints extend that comparison without replacing the independent two-task protocol below.
 **Descriptive results; contamination scan of the training lineage against these tasks is pending.** Belebele and MMLU-PL are near their four-choice 25% random baseline. No overall average, ranking or release threshold is inferred. Numbers from different evaluators must not be silently substituted: prompting, BOS handling, normalization and software versions can change the result.
 There is no replay-only control for this continuation, and initialization, exposure and schedule differ between checkpoints. These measurements do not isolate a causal effect of the new datasets or quantify training-seed variance. Earlier internal diagnostics are separate historical records, not a criterion for naming or publishing this model.
 ## Additional GSM8K and SST-2 measurements — update 10,000
+SlayerLab measured this two-task subset on 11 November 2026 using the [pinned BenchTeam code](https://github.com/slayerlabs/benchmarks-fabryka-ai/tree/8b547007526971bd2fc35dfe56ff2c7d9b802568), lm-eval 0.4.9.1, the `slayer` adapter, float32, batch 1 and context limit 4,096. These are full evaluated splits, not a full 50-task run or a copy of the community's independent report. **This separate two-task protocol measures 1.09 update 10,000 only; it does not supply GSM8K or SST-2 results for updates 15,000 or 20,000.** Update 10,000 is an internal working checkpoint, not a published or inference-ready 1.09 weight release.
 | Task | Metric / prompting | n | 1.07 | 1.08 | 1.09, update 10,000 |
 |---|---|---:|---:|---:|---:|
 ## Cost, licensing and limitations
+The total campaign ceiling remains **USD 425**, including GPU allocations, preparation, interruptions, storage and evaluation tracked by the resource meter. At the operator's 11 November, 10:39 UTC snapshot, the meter reported **approximately USD 239 spent**, leaving approximately **USD 186**; this is counter attribution, not a settled invoice or a permanently current balance. Hardware selection is based on effective cost per processed position and verified availability. The next budget-dependent length review is at update 30,000, before the selected endpoint's WSD decay boundary. No higher spending limit or guaranteed completion date is implied.
 The selected 1.08 base declares Apache-2.0; source-text rights and future publication still require separate approval. Privacy filters are heuristic; generated text may be inaccurate, biased, offensive or personal. No reasoning quality, instruction-following ability or professional reliability has been demonstrated for the revised 1.09.