Upload training_logs/it_phase57_partB_h8_bare_multidistance_t8__seed42.json with huggingface_hub

training_logs/it_phase57_partB_h8_bare_multidistance_t8__seed42.json

@@ -0,0 +1,98 @@
+{
+  "loss": [
+    192895.625,
+    83181.578125,
+    75872.6953125,
+    72378.8203125,
+    69673.5390625,
+    68143.6640625,
+    67047.8359375,
+    66126.4921875,
+    65826.4375,
+    65016.96875,
+    64892.83984375,
+    64677.31640625,
+    64578.00390625,
+    64231.3359375,
+    63977.31640625,
+    63754.2734375,
+    63827.35546875
+  ],
+  "l0": [
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0,
+    500.0
+  ],
+  "steps_logged": [
+    0,
+    200,
+    400,
+    600,
+    800,
+    1000,
+    1200,
+    1400,
+    1600,
+    1800,
+    2000,
+    2200,
+    2400,
+    2600,
+    2800,
+    3000,
+    3200
+  ],
+  "final_step": 3200,
+  "converged": true,
+  "plateau_last": 0.018900497634682014,
+  "elapsed_s": 12054.72945189476,
+  "shifts": [
+    1,
+    2,
+    4
+  ],
+  "matryoshka_h_size": 3686,
+  "alpha": null,
+  "row": 35,
+  "arch_id": "phase57_partB_h8_bare_multidistance_t8",
+  "arch": "phase57_partB_h8_bare_multidistance_t8",
+  "group": 4,
+  "src_class": "TXCBareMultiDistanceContrastiveAntidead",
+  "src_module": "src.architectures.txc_bare_multidistance_contrastive_antidead",
+  "T": 8,
+  "T_max": null,
+  "t_sample": null,
+  "n_layers": null,
+  "k_win": 500,
+  "k_pos": 62,
+  "gamma": null,
+  "n_scales": null,
+  "seed": 42,
+  "d_in": 2304,
+  "d_sae": 18432,
+  "subject_model": "google/gemma-2-2b-it",
+  "anchor_layer": 13,
+  "mlc_layers": [
+    11,
+    12,
+    13,
+    14,
+    15
+  ],
+  "phase": "phase7_unification",
+  "run_id": "it_phase57_partB_h8_bare_multidistance_t8__seed42"
+}