han1823123123 commited on
Commit
f9dbeb8
·
verified ·
1 Parent(s): bf7b693

Upload folder using huggingface_hub

Browse files
05b5e0096877cd3f/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "act_cache_key": "f3ddc5b02afbe4b6",
3
+ "agent": "unknown",
4
+ "arch": "tsae_paper",
5
+ "arch_version": "1.0.0",
6
+ "datasource": "toy_markov_n20_d40",
7
+ "saved_ts": "2026-05-05T21:19:19Z",
8
+ "seed": 1,
9
+ "train_key": "05b5e0096877cd3f",
10
+ "training_cfg": {
11
+ "arch_hparams_override": {
12
+ "k_pos": 5
13
+ },
14
+ "batch_size": 1024,
15
+ "bricken_enabled": false,
16
+ "bricken_max_resample_fraction": 0.5,
17
+ "bricken_min_fires": 1,
18
+ "bricken_n_check": 2048,
19
+ "bricken_resample_every": 500,
20
+ "dead_threshold_tokens": 10000000,
21
+ "ema_auxk_alpha": 0.03125,
22
+ "learning_rate": 0.0003,
23
+ "n_steps": 30000,
24
+ "optimizer": "adam",
25
+ "plateau_early_stop": false,
26
+ "plateau_min_delta": 0.0001,
27
+ "plateau_patience": 5000,
28
+ "precision": "bf16",
29
+ "train_window_size": null,
30
+ "warmup_steps": 1000
31
+ }
32
+ }
05b5e0096877cd3f/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f6eb4bc3e567936b0c311be2f7ed9cde71103dda85450c5d67335431323a793b
3
+ size 14000