Upload batch 41/100 (1000 files)

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-1675/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8d425312f062a6d24e44b40e74911448cb5d3d5f5c65877e11c570decf27129
+size 39641

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-1675/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.174,
+    "mse": 0.061786928011359904,
+    "corr_pred_acc": 0.7107337317694175,
+    "corr_pred_cons": 0.5243444211308688,
+    "corr_cons_acc": 0.44940073047931717
+  },
+  "test": {
+    "0.05": 0.24,
+    "mse": 0.06550736061454351,
+    "corr_pred_acc": 0.7761215337068688,
+    "corr_pred_cons": 0.6538478879935853,
+    "corr_cons_acc": 0.5433606127799882
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-1675/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 28,
+  "n_layers_used": 28,
+  "hidden_size": 3584,
+  "model": "Qwen2.5-7B-Instruct",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-2144/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3abdcc78a9096ca1be5666a122603d12b7329b3bc3d8c7ca55fe4857391c0923
+size 39577

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-2144/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.261,
+    "mse": 0.049692810073484764,
+    "corr_pred_acc": 0.7735607313189128,
+    "corr_pred_cons": 0.5028305751390535,
+    "corr_cons_acc": 0.44940073047931717
+  },
+  "test": {
+    "0.05": 0.322,
+    "mse": 0.05202159053758131,
+    "corr_pred_acc": 0.8276294942297087,
+    "corr_pred_cons": 0.6297444237480252,
+    "corr_cons_acc": 0.5433606127799882
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-2144/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 28,
+  "n_layers_used": 28,
+  "hidden_size": 3584,
+  "model": "Qwen2.5-7B-Instruct",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-3685/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a8a28217e89ac1f464d4aa8835f25703ea9314a46af87505f325beaa6f39129
+size 39577

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-3685/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.364,
+    "mse": 0.0338395499187624,
+    "corr_pred_acc": 0.8645623894292904,
+    "corr_pred_cons": 0.47084650915859,
+    "corr_cons_acc": 0.44940073047931717
+  },
+  "test": {
+    "0.05": 0.468,
+    "mse": 0.034437587095792095,
+    "corr_pred_acc": 0.8895887081125468,
+    "corr_pred_cons": 0.5795833763659817,
+    "corr_cons_acc": 0.5433606127799882
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-3685/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 28,
+  "n_layers_used": 28,
+  "hidden_size": 3584,
+  "model": "Qwen2.5-7B-Instruct",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-3819/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6074698d43c2b3ee43aeabd462ae4705ec24a0d183722e1f90f6b22ca6f7efef
+size 39577

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-3819/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.429,
+    "mse": 0.029710313330236056,
+    "corr_pred_acc": 0.8755064080514924,
+    "corr_pred_cons": 0.4713766503890678,
+    "corr_cons_acc": 0.44940073047931717
+  },
+  "test": {
+    "0.05": 0.507,
+    "mse": 0.0322717165316233,
+    "corr_pred_acc": 0.8967361740648405,
+    "corr_pred_cons": 0.5949894152082011,
+    "corr_cons_acc": 0.5433606127799882
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-3819/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 28,
+  "n_layers_used": 28,
+  "hidden_size": 3584,
+  "model": "Qwen2.5-7B-Instruct",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-4154/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57deb689ae12e9dbafa7628c19117da3afaeeef65c53424367f02a35660a7fc9
+size 39513

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-4154/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.449,
+    "mse": 0.029215179059417416,
+    "corr_pred_acc": 0.878396776707274,
+    "corr_pred_cons": 0.4895903577509612,
+    "corr_cons_acc": 0.44940073047931717
+  },
+  "test": {
+    "0.05": 0.547,
+    "mse": 0.030632813645972236,
+    "corr_pred_acc": 0.9033470656497372,
+    "corr_pred_cons": 0.572540395941644,
+    "corr_cons_acc": 0.5433606127799882
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-4154/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 28,
+  "n_layers_used": 28,
+  "hidden_size": 3584,
+  "model": "Qwen2.5-7B-Instruct",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-4288/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:160678a551b470a9016f4b89ff4380eeac8a6346aa7977a36db473ad1213321e
+size 39641

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-4288/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.472,
+    "mse": 0.026849055478002096,
+    "corr_pred_acc": 0.8896081208081348,
+    "corr_pred_cons": 0.47806340785300805,
+    "corr_cons_acc": 0.44940073047931717
+  },
+  "test": {
+    "0.05": 0.545,
+    "mse": 0.03587788504491612,
+    "corr_pred_acc": 0.884927402746559,
+    "corr_pred_cons": 0.5804529871522189,
+    "corr_cons_acc": 0.5433606127799882
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-4288/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 28,
+  "n_layers_used": 28,
+  "hidden_size": 3584,
+  "model": "Qwen2.5-7B-Instruct",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-5494/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c987f86c8e702f670b3b0dced78fef4b8c730d6e8c0129161988382898125063
+size 39513

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-5494/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.539,
+    "mse": 0.01769074466441364,
+    "corr_pred_acc": 0.9256926111276811,
+    "corr_pred_cons": 0.4687793953909748,
+    "corr_cons_acc": 0.44940073047931717
+  },
+  "test": {
+    "0.05": 0.629,
+    "mse": 0.023019853707382406,
+    "corr_pred_acc": 0.9279045010458388,
+    "corr_pred_cons": 0.5631092546722352,
+    "corr_cons_acc": 0.5433606127799882
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-5494/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 28,
+  "n_layers_used": 28,
+  "hidden_size": 3584,
+  "model": "Qwen2.5-7B-Instruct",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-5561/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce31fdc0be52626a3d558ca1879f16905d32f6c8e180f697597b28e3d4db76d3
+size 39449

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-5561/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.524,
+    "mse": 0.018265177129132667,
+    "corr_pred_acc": 0.9231912621840672,
+    "corr_pred_cons": 0.44810209856835587,
+    "corr_cons_acc": 0.44940073047931717
+  },
+  "test": {
+    "0.05": 0.625,
+    "mse": 0.02455975216014587,
+    "corr_pred_acc": 0.9250361917284045,
+    "corr_pred_cons": 0.538338652114522,
+    "corr_cons_acc": 0.5433606127799882
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-5561/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 28,
+  "n_layers_used": 28,
+  "hidden_size": 3584,
+  "model": "Qwen2.5-7B-Instruct",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-603/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbfe88005ea6ff4568b06584103a791cbb9221e673c82003c0b23a7f83eb3a14
+size 39833

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-603/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.146,
+    "mse": 0.07475311160663348,
+    "corr_pred_acc": 0.6484608001627474,
+    "corr_pred_cons": 0.5087648397567388,
+    "corr_cons_acc": 0.44940073047931717
+  },
+  "test": {
+    "0.05": 0.105,
+    "mse": 0.08856549134199711,
+    "corr_pred_acc": 0.7144452569304933,
+    "corr_pred_cons": 0.6403626027259307,
+    "corr_cons_acc": 0.5433606127799882
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-603/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 28,
+  "n_layers_used": 28,
+  "hidden_size": 3584,
+  "model": "Qwen2.5-7B-Instruct",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-871/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d4947fbe236fd5c3926c82477c0a5ff8e65a6113fba53c68aaaf0c61ffce7b3
+size 39321

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-871/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.089,
+    "mse": 0.14946001203746023,
+    "corr_pred_acc": 0.4520385298235616,
+    "corr_pred_cons": 0.44429549671614826,
+    "corr_cons_acc": 0.44940073047931717
+  },
+  "test": {
+    "0.05": 0.197,
+    "mse": 0.29895331224981864,
+    "corr_pred_acc": 0.534932755624288,
+    "corr_pred_cons": 0.5490952745374281,
+    "corr_cons_acc": 0.5433606127799882
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen2.5-7B-Instruct_1_5e-05_0.01_8_8_0.1_probe-end_cca2eb6c/checkpoint-871/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 28,
+  "n_layers_used": 28,
+  "hidden_size": 3584,
+  "model": "Qwen2.5-7B-Instruct",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-0/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3bd983319ce610426328863f6e199d6aadf6af37b263a8096b99783770ccefdc
+size 40473

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-0/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.086,
+    "mse": 0.07862197410952848,
+    "corr_pred_acc": 0.33790394763469234,
+    "corr_pred_cons": 0.2993155151968401,
+    "corr_cons_acc": 0.8778464815290765
+  },
+  "test": {
+    "0.05": 0.115,
+    "mse": 0.08134418814635157,
+    "corr_pred_acc": 0.4464617626692198,
+    "corr_pred_cons": 0.4900580746231971,
+    "corr_cons_acc": 0.8939567786442054
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-0/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 32,
+  "n_layers_used": 32,
+  "hidden_size": 4096,
+  "model": "Qwen3.5-9B-Base",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1005/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b0a48e6cefd96f28934ed1eb3571f53f1d32c9420524a399e9aa503ba7eba9c
+size 39897

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1005/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.685,
+    "mse": 0.004509785237326687,
+    "corr_pred_acc": 0.9756041651110294,
+    "corr_pred_cons": 0.8671528967700693,
+    "corr_cons_acc": 0.8778464815290765
+  },
+  "test": {
+    "0.05": 0.737,
+    "mse": 0.0031909738863797744,
+    "corr_pred_acc": 0.9826351429095547,
+    "corr_pred_cons": 0.8621581858631515,
+    "corr_cons_acc": 0.8939567786442054
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1005/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 32,
+  "n_layers_used": 32,
+  "hidden_size": 4096,
+  "model": "Qwen3.5-9B-Base",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1072/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc1dc0b415ba2d0737325f4fa84eb8e75624ba92301b1877504f8265d6634b9e
+size 39769

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1072/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.684,
+    "mse": 0.0034771493954820513,
+    "corr_pred_acc": 0.9798592807853214,
+    "corr_pred_cons": 0.8567135712803141,
+    "corr_cons_acc": 0.8778464815290765
+  },
+  "test": {
+    "0.05": 0.689,
+    "mse": 0.004402357435603943,
+    "corr_pred_acc": 0.9769943918275018,
+    "corr_pred_cons": 0.8408829484310651,
+    "corr_cons_acc": 0.8939567786442054
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1072/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 32,
+  "n_layers_used": 32,
+  "hidden_size": 4096,
+  "model": "Qwen3.5-9B-Base",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1139/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f7a32180700cfb1b630a358a8e8b5410067259cc52b0b57447e9620ce1bb3dd1
+size 39577

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1139/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.612,
+    "mse": 0.006427536765477978,
+    "corr_pred_acc": 0.9734806113891177,
+    "corr_pred_cons": 0.8284661364261018,
+    "corr_cons_acc": 0.8778464815290765
+  },
+  "test": {
+    "0.05": 0.635,
+    "mse": 0.006652601694456758,
+    "corr_pred_acc": 0.9706597927921877,
+    "corr_pred_cons": 0.811766750360122,
+    "corr_cons_acc": 0.8939567786442054
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1139/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 32,
+  "n_layers_used": 32,
+  "hidden_size": 4096,
+  "model": "Qwen3.5-9B-Base",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1206/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67e534d6207565feee8ebcefa2d940e3942908dd80184437edffc945d927120f
+size 39769

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1206/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.722,
+    "mse": 0.004057810282258648,
+    "corr_pred_acc": 0.9786011419488293,
+    "corr_pred_cons": 0.8508729652177204,
+    "corr_cons_acc": 0.8778464815290765
+  },
+  "test": {
+    "0.05": 0.798,
+    "mse": 0.0023846673576462737,
+    "corr_pred_acc": 0.9861105006330337,
+    "corr_pred_cons": 0.8708820205919977,
+    "corr_cons_acc": 0.8939567786442054
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1206/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 32,
+  "n_layers_used": 32,
+  "hidden_size": 4096,
+  "model": "Qwen3.5-9B-Base",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1273/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef51caa9fd506b1b18d6e12bfd58ead56464fe4899b05ac080e6687e87b30f60
+size 39833

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1273/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.647,
+    "mse": 0.006753641949375595,
+    "corr_pred_acc": 0.9705775444511159,
+    "corr_pred_cons": 0.8785122334060139,
+    "corr_cons_acc": 0.8778464815290765
+  },
+  "test": {
+    "0.05": 0.702,
+    "mse": 0.003656907859341011,
+    "corr_pred_acc": 0.9800871866659424,
+    "corr_pred_cons": 0.8520497010553253,
+    "corr_cons_acc": 0.8939567786442054
+  }
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-1273/probe_meta.json

@@ -0,0 +1,9 @@
+{
+  "probe": "end",
+  "n_layers_full": 32,
+  "n_layers_used": 32,
+  "hidden_size": 4096,
+  "model": "Qwen3.5-9B-Base",
+  "dataset": "math10_categoriesbalanced",
+  "lora_rank": 8
+}

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-134/evals/math10_categoriesbalanced.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:776923974acb32d1bb9339b4e0cac7ad9923aaf722d017c9b58fb361fbc6f366
+size 39065

models/default/unsloth_full/math10_categoriesbalanced_Qwen3.5-9B-Base_1_5e-05_0.01_8_8_0.1_probe-end_50fa036c/checkpoint-134/evals/math10_categoriesbalanced_metrics.json

@@ -0,0 +1,16 @@
+{
+  "train_subset": {
+    "0.05": 0.125,
+    "mse": 0.07721839789032262,
+    "corr_pred_acc": 0.6586269237147787,
+    "corr_pred_cons": 0.6378467327280705,
+    "corr_cons_acc": 0.8778464815290765
+  },
+  "test": {
+    "0.05": 0.132,
+    "mse": 0.07488967621934565,
+    "corr_pred_acc": 0.7832613862967734,
+    "corr_pred_cons": 0.8436222210016883,
+    "corr_cons_acc": 0.8939567786442054
+  }
+}