seed: stage1_v2/stage1_summary.json
Browse files
stage1_v2/stage1_summary.json
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model": "Qwen/Qwen3.5-0.8B-Base",
|
| 3 |
+
"probe": "probe_results/qwen3_5_0_8b_base_v2.json",
|
| 4 |
+
"steps": 300,
|
| 5 |
+
"lr": 0.001,
|
| 6 |
+
"seq_len": 8192,
|
| 7 |
+
"batch_size": 1,
|
| 8 |
+
"d_idx": 16,
|
| 9 |
+
"elapsed_min": 3.791798701449928,
|
| 10 |
+
"final_per_head_kl": {
|
| 11 |
+
"7": [
|
| 12 |
+
1.584153175354004
|
| 13 |
+
],
|
| 14 |
+
"11": [
|
| 15 |
+
1.064223051071167
|
| 16 |
+
],
|
| 17 |
+
"15": [
|
| 18 |
+
1.0976929664611816,
|
| 19 |
+
1.2519683837890625,
|
| 20 |
+
1.1221760511398315,
|
| 21 |
+
0.981181263923645,
|
| 22 |
+
1.2290284633636475
|
| 23 |
+
]
|
| 24 |
+
},
|
| 25 |
+
"final_mean_kl": 1.1900604793003626,
|
| 26 |
+
"final_max_kl": 1.584153175354004
|
| 27 |
+
}
|