Upload training_logs/phase57_partB_h8_bare_multidistance_t10__seed42.json with huggingface_hub

training_logs/phase57_partB_h8_bare_multidistance_t10__seed42.json

@@ -0,0 +1,104 @@
+{
+  "loss": [
+    105003.7109375,
+    52358.4921875,
+    47925.06640625,
+    45527.77734375,
+    44065.27734375,
+    43462.57421875,
+    42782.03515625,
+    42197.1640625,
+    41981.99609375,
+    41768.98046875,
+    41431.15234375,
+    41320.86328125,
+    41095.10546875,
+    41000.9765625,
+    40877.00390625,
+    40746.453125,
+    40753.51171875,
+    40709.9140625
+  ],
+  "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,
+    500.0
+  ],
+  "steps_logged": [
+    0,
+    200,
+    400,
+    600,
+    800,
+    1000,
+    1200,
+    1400,
+    1600,
+    1800,
+    2000,
+    2200,
+    2400,
+    2600,
+    2800,
+    3000,
+    3200,
+    3400
+  ],
+  "final_step": 3400,
+  "converged": true,
+  "plateau_last": 0.016908817185128506,
+  "elapsed_s": 24616.253779649734,
+  "T": 10,
+  "shifts": [
+    1,
+    2,
+    5
+  ],
+  "matryoshka_h_size": 3686,
+  "alpha": 1.0,
+  "src_class": "TXCBareMultiDistanceContrastiveAntidead",
+  "n_seqs_used": 24000,
+  "ctx_used": 64,
+  "ctx_slice_direction": "last",
+  "row": 37,
+  "arch_id": "phase57_partB_h8_bare_multidistance_t10",
+  "arch": "phase57_partB_h8_bare_multidistance_t10",
+  "group": 4,
+  "src_module": "src.architectures.txc_bare_multidistance_contrastive_antidead",
+  "T_max": null,
+  "t_sample": null,
+  "n_layers": null,
+  "k_win": 500,
+  "k_pos": 50,
+  "gamma": null,
+  "n_scales": null,
+  "seed": 42,
+  "d_in": 2304,
+  "d_sae": 18432,
+  "subject_model": "google/gemma-2-2b",
+  "anchor_layer": 12,
+  "mlc_layers": [
+    10,
+    11,
+    12,
+    13,
+    14
+  ],
+  "phase": "phase7_unification",
+  "run_id": "phase57_partB_h8_bare_multidistance_t10__seed42"
+}