ToluClassics commited on
Commit
6b2637b
1 Parent(s): 2035879
Files changed (6) hide show
  1. config.json +37 -0
  2. optimizer.pt +3 -0
  3. pytorch_model.bin +3 -0
  4. rng_state.pth +3 -0
  5. scheduler.pt +3 -0
  6. trainer_state.json +0 -0
config.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/t5-v1_1-base",
3
+ "architectures": [
4
+ "T5EncoderModel"
5
+ ],
6
+ "d_ff": 2048,
7
+ "d_kv": 64,
8
+ "d_model": 768,
9
+ "decoder_start_token_id": 0,
10
+ "dense_act_fn": "gelu_new",
11
+ "dropout_rate": 0.1,
12
+ "eos_token_id": 1,
13
+ "feed_forward_proj": "gated-gelu",
14
+ "id2label": {
15
+ "0": "LABEL_0"
16
+ },
17
+ "initializer_factor": 1.0,
18
+ "is_encoder_decoder": true,
19
+ "is_gated_act": true,
20
+ "label2id": {
21
+ "LABEL_0": 0
22
+ },
23
+ "layer_norm_epsilon": 1e-06,
24
+ "model_type": "t5",
25
+ "num_decoder_layers": 12,
26
+ "num_heads": 12,
27
+ "num_layers": 12,
28
+ "output_past": true,
29
+ "pad_token_id": 0,
30
+ "relative_attention_max_distance": 128,
31
+ "relative_attention_num_buckets": 32,
32
+ "tie_word_embeddings": false,
33
+ "torch_dtype": "float32",
34
+ "transformers_version": "4.26.0",
35
+ "use_cache": true,
36
+ "vocab_size": 32128
37
+ }
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc5a3bbc6ca57939c07af473c13add64c2552362008c2e0d1fa623eebd024a7d
3
+ size 996037
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d8d71303df9fb4a237b2cede9776d61cd199961c174bf2401ef7d676d6307a0b
3
+ size 438552265
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fde4c40cbf59d81c61a896520958f31614c88793d6de606e7279f9774d50903
3
+ size 14575
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aab76a59e40ba0fb6d7718d1fe4d858167cddf9b60db3472f3123cb7729fdc08
3
+ size 627
trainer_state.json ADDED
The diff for this file is too large to render. See raw diff