Hugging Face's logo

phonemetransformers
/
babylm-small-gpt2_lm-model

Safetensors
gpt2
Model card Files Community
codebyzeb commited on
Commit
40b2009
•
1 Parent(s): e52af17

Training in progress, step 250000

Browse files
Files changed (5) hide show
  1. babyslm/syntactic.txt +0 -0
  2. blimp_results.json +2 -2
  3. config.json +3 -3
  4. model.safetensors +1 -1
  5. training_args.bin +1 -1
babyslm/syntactic.txt CHANGED
The diff for this file is too large to render. See raw diff
 
blimp_results.json CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ee9f8e9be1a551d53ec2c6030ee83241c07d47f0de3037516116c0671cfc0dfa
3
- size 80222748
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bfea65bfcce37100d270202f3ad83c790c61a8aaf8ae43bbab265110bd03fd1b
3
+ size 80268261
config.json CHANGED
@@ -3,9 +3,9 @@
3
  "architectures": [
4
  "GPT2LMHeadModel"
5
  ],
6
- "attn_pdrop": 0.3,
7
  "bos_token_id": 0,
8
- "embd_pdrop": 0.3,
9
  "eos_token_id": 0,
10
  "initializer_range": 0.02,
11
  "layer_norm_epsilon": 1e-05,
@@ -16,7 +16,7 @@
16
  "n_layer": 12,
17
  "n_positions": 256,
18
  "reorder_and_upcast_attn": false,
19
- "resid_pdrop": 0.3,
20
  "scale_attn_by_inverse_layer_idx": false,
21
  "scale_attn_weights": true,
22
  "summary_activation": null,
 
3
  "architectures": [
4
  "GPT2LMHeadModel"
5
  ],
6
+ "attn_pdrop": 0.1,
7
  "bos_token_id": 0,
8
+ "embd_pdrop": 0.1,
9
  "eos_token_id": 0,
10
  "initializer_range": 0.02,
11
  "layer_norm_epsilon": 1e-05,
 
16
  "n_layer": 12,
17
  "n_positions": 256,
18
  "reorder_and_upcast_attn": false,
19
+ "resid_pdrop": 0.1,
20
  "scale_attn_by_inverse_layer_idx": false,
21
  "scale_attn_weights": true,
22
  "summary_activation": null,
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:55ffbd17507860bdc45e73c5738e28e9f9d1d04a0db7dfdd62d509773786ee76
3
  size 390177408
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bb924aeafeef2682cf5636ae4727d725b15812a11402f8efa7fc46da860de76f
3
  size 390177408
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b6bf434d08adffb63320a056f43c3d312c362758a72db108fcc5c86bb963bff1
3
  size 5368
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ae0dde72c52f7e1e0ff9bbc8aac95f1da19222bc12c580df46c08d305ad1f34
3
  size 5368