han1823123123 commited on
Commit
cb9600a
·
verified ·
1 Parent(s): c8ef67b

agent=agent_steer_100k train_key=c85e7c1b4ab4900b

Browse files
c85e7c1b4ab4900b/config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "act_cache_key": "f01ca87f2e8f3365",
3
+ "agent": "agent_steer_100k",
4
+ "arch": "tsae_paper",
5
+ "arch_version": "1.0.0",
6
+ "datasource": "gemma_2_2b_base_l13_fineweb_24k128",
7
+ "saved_ts": "2026-05-05T18:38:32Z",
8
+ "seed": 2,
9
+ "train_key": "c85e7c1b4ab4900b",
10
+ "training_cfg": {
11
+ "arch_hparams_override": null,
12
+ "batch_size": 1024,
13
+ "bricken_enabled": false,
14
+ "bricken_max_resample_fraction": 0.5,
15
+ "bricken_min_fires": 1,
16
+ "bricken_n_check": 2048,
17
+ "bricken_resample_every": 500,
18
+ "dead_threshold_tokens": 10000000,
19
+ "ema_auxk_alpha": 0.03125,
20
+ "learning_rate": 0.0003,
21
+ "n_steps": 20000,
22
+ "optimizer": "adam",
23
+ "plateau_early_stop": false,
24
+ "plateau_min_delta": 0.0001,
25
+ "plateau_patience": 5000,
26
+ "precision": "bf16",
27
+ "train_window_size": 2,
28
+ "warmup_steps": 1000
29
+ }
30
+ }
c85e7c1b4ab4900b/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f182fff4e3d02d73b079c7a4a8c332c91ebfbe286361fb8c620d0aef3ae7aba
3
+ size 302196344