AbstractPhil commited on
Commit
22611d4
·
verified ·
1 Parent(s): 9224b9c

Upload folder using huggingface_hub

Browse files
Files changed (4) hide show
  1. beeper_final.pt +3 -0
  2. beeper_final.safetensors +3 -0
  3. config.json +47 -0
  4. tokenizer.json +0 -0
beeper_final.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:053d16f81c5f0091fedc994f6901c890291df9c5ce8e0f102fdf73fd7609150b
3
+ size 110234924
beeper_final.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4148b1b7b7646abec817f34d1da0349068540de25b0efa75979049b067a5691
3
+ size 110221080
config.json ADDED
@@ -0,0 +1,47 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "Beeper",
3
+ "dataset_name": "roneneldan/TinyStories",
4
+ "dataset_split": "train[:10%]",
5
+ "context": 512,
6
+ "vocab_size": 8192,
7
+ "add_bos_eos": true,
8
+ "val_ratio": 0.01,
9
+ "test_ratio": 0.01,
10
+ "dim": 512,
11
+ "n_layers": 6,
12
+ "n_heads": 8,
13
+ "mlp_ratio": 4.0,
14
+ "dropout": 0.0,
15
+ "resid_dropout": 0.1,
16
+ "grad_checkpoint": false,
17
+ "compile_model": true,
18
+ "batch_size": 32,
19
+ "grad_accum_steps": 1,
20
+ "epochs": 3,
21
+ "lr": 0.0003,
22
+ "betas": [
23
+ 0.9,
24
+ 0.95
25
+ ],
26
+ "weight_decay": 0.1,
27
+ "warmup_steps": 500,
28
+ "max_steps": null,
29
+ "clip_grad": 1.0,
30
+ "min_lr": 1e-06,
31
+ "label_smoothing": 0.0,
32
+ "mixed_precision": "bf16",
33
+ "log_dir": "./runs/beeper",
34
+ "log_interval": 50,
35
+ "ckpt_dir": "./beeper_checkpoints",
36
+ "export_dir": "./beeper_export",
37
+ "temperature": 0.9,
38
+ "top_k": 40,
39
+ "top_p": 0.9,
40
+ "repetition_penalty": 1.1,
41
+ "presence_penalty": 0.6,
42
+ "frequency_penalty": 0.0,
43
+ "upload_to_hub": true,
44
+ "hf_repo": "AbstractPhil/beeper-tinystories-6l-512d-ctx512",
45
+ "tokenizer_path": "beeper.tokenizer.json",
46
+ "seed": 1337
47
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff