Update experiment measurements and status

reports/mixtures/control/manifest.json

@@ -0,0 +1,8 @@
+{
+  "tokens": 524288000,
+  "sha256": "95a5780cd191654957308b97d36bb1e4dd22166e0be05b9ec2d97292b54715e8",
+  "source": "control",
+  "replacement_fraction": 0,
+  "common_base_windows_seed": 4401,
+  "note": "Physical token fractions; randomized training window offsets can straddle source boundaries."
+}

reports/mixtures/cosmopedia/manifest.json

@@ -0,0 +1,8 @@
+{
+  "tokens": 524288000,
+  "sha256": "1b762667a92f1f9fd9d4172030cc0f7c0a7648e1ce663fc5da19648e4eb973bd",
+  "source": "cosmopedia",
+  "replacement_fraction": 0.2,
+  "common_base_windows_seed": 4401,
+  "note": "Physical token fractions; randomized training window offsets can straddle source boundaries."
+}

reports/mixtures/dclm_edu/manifest.json

@@ -0,0 +1,8 @@
+{
+  "tokens": 524288000,
+  "sha256": "5636016c21ba4b60a9cd1112b993063fe9c14f8c214db3be540d955e6015172c",
+  "source": "dclm_edu",
+  "replacement_fraction": 0.2,
+  "common_base_windows_seed": 4401,
+  "note": "Physical token fractions; randomized training window offsets can straddle source boundaries."
+}

reports/mixtures/finemath/manifest.json

@@ -0,0 +1,8 @@
+{
+  "tokens": 524288000,
+  "sha256": "e8fc6860eb6a3f0cad495835c87048a9d8b0df980d85def904dc01742b5f100f",
+  "source": "finemath",
+  "replacement_fraction": 0.2,
+  "common_base_windows_seed": 4401,
+  "note": "Physical token fractions; randomized training window offsets can straddle source boundaries."
+}

reports/mixtures/fineweb_edu/manifest.json

@@ -0,0 +1,8 @@
+{
+  "tokens": 524288000,
+  "sha256": "c05736407e0b1730dce6da63e063e1fe092736404dc6b2a0247a943037058571",
+  "source": "fineweb_edu",
+  "replacement_fraction": 0.2,
+  "common_base_windows_seed": 4401,
+  "note": "Physical token fractions; randomized training window offsets can straddle source boundaries."
+}

reports/sources/fineweb_edu/manifest.json

@@ -0,0 +1,34 @@
+{
+  "repo": "HuggingFaceFW/fineweb-edu",
+  "revision": "87f09149ef4734204d70ed1d046ddc9ca3f2b8f9",
+  "files": [
+    "sample/10BT/000_00000.parquet",
+    "sample/10BT/001_00000.parquet",
+    "sample/10BT/002_00000.parquet",
+    "sample/10BT/003_00000.parquet",
+    "sample/10BT/004_00000.parquet",
+    "sample/10BT/005_00000.parquet",
+    "sample/10BT/006_00000.parquet",
+    "sample/10BT/007_00000.parquet",
+    "sample/10BT/008_00000.parquet",
+    "sample/10BT/009_00000.parquet",
+    "sample/10BT/010_00000.parquet",
+    "sample/10BT/011_00000.parquet",
+    "sample/10BT/012_00000.parquet",
+    "sample/10BT/013_00000.parquet"
+  ],
+  "license": "odc-by",
+  "used_files": [
+    {
+      "file": "sample/10BT/007_00000.parquet",
+      "sha256": "4b7d1f697d4afff7f5f65cb7aa4d83acc6db716da4d46ea23399d0c50608c6e3"
+    }
+  ],
+  "train_tokens": 120040261,
+  "validation_tokens": 203096,
+  "accepted_documents": 102831,
+  "rejected_overlap": 53,
+  "train_sha256": "ea51d6566f76ec7d59edbdc688dc2c2a81d34e0ed29dd8b17ff61398ff86f3d1",
+  "validation_sha256": "e05b4fdf2b4332d0bc8f87fd01f004148554815f7fc4509c3e6926286d6cee43",
+  "audit": "normalized 13-word exact overlap plus full short sentences; not semantic decontamination; legacy ARC-MIX was not re-audited here"
+}