Upload FSMTForConditionalGeneration

#5
Files changed (4) hide show
  1. README.md +1 -1
  2. config.json +32 -22
  3. generation_config.json +1 -1
  4. model.safetensors +2 -2
README.md CHANGED
@@ -1,5 +1,5 @@
1
  ---
2
- language:
3
  - de
4
  - en
5
  tags:
 
1
  ---
2
+ language:
3
  - de
4
  - en
5
  tags:
config.json CHANGED
@@ -1,37 +1,47 @@
1
  {
 
 
2
  "architectures": [
3
  "FSMTForConditionalGeneration"
4
  ],
5
- "model_type": "fsmt",
6
- "activation_dropout": 0.0,
7
- "activation_function": "relu",
8
  "attention_dropout": 0.1,
 
9
  "d_model": 1024,
 
 
 
 
 
 
 
 
 
 
10
  "dropout": 0.2,
11
- "init_std": 0.02,
12
- "max_position_embeddings": 1024,
13
- "num_hidden_layers": 6,
14
- "src_vocab_size": 42024,
15
- "tgt_vocab_size": 42024,
16
- "langs": [
17
- "de",
18
- "en"
19
- ],
20
  "encoder_attention_heads": 16,
21
  "encoder_ffn_dim": 8192,
22
  "encoder_layerdrop": 0,
23
  "encoder_layers": 6,
24
- "decoder_attention_heads": 16,
25
- "decoder_ffn_dim": 4096,
26
- "decoder_layerdrop": 0,
27
- "decoder_layers": 6,
28
- "bos_token_id": 0,
29
- "pad_token_id": 1,
30
  "eos_token_id": 2,
 
 
31
  "is_encoder_decoder": true,
 
 
 
 
 
 
 
 
 
 
 
32
  "scale_embedding": true,
 
 
33
  "tie_word_embeddings": true,
34
- "num_beams": 5,
35
- "early_stopping": false,
36
- "length_penalty": 1.1
37
- }
 
1
  {
2
+ "activation_dropout": 0.0,
3
+ "activation_function": "relu",
4
  "architectures": [
5
  "FSMTForConditionalGeneration"
6
  ],
 
 
 
7
  "attention_dropout": 0.1,
8
+ "bos_token_id": 0,
9
  "d_model": 1024,
10
+ "decoder": {
11
+ "bos_token_id": 2,
12
+ "model_type": "fsmt_decoder",
13
+ "vocab_size": 42024
14
+ },
15
+ "decoder_attention_heads": 16,
16
+ "decoder_ffn_dim": 4096,
17
+ "decoder_layerdrop": 0,
18
+ "decoder_layers": 6,
19
+ "decoder_start_token_id": 2,
20
  "dropout": 0.2,
 
 
 
 
 
 
 
 
 
21
  "encoder_attention_heads": 16,
22
  "encoder_ffn_dim": 8192,
23
  "encoder_layerdrop": 0,
24
  "encoder_layers": 6,
 
 
 
 
 
 
25
  "eos_token_id": 2,
26
+ "forced_eos_token_id": 2,
27
+ "init_std": 0.02,
28
  "is_encoder_decoder": true,
29
+ "langs": [
30
+ "de",
31
+ "en"
32
+ ],
33
+ "length_penalty": null,
34
+ "max_length": 200,
35
+ "max_position_embeddings": 1024,
36
+ "model_type": "fsmt",
37
+ "num_beams": 5,
38
+ "num_hidden_layers": 6,
39
+ "pad_token_id": 1,
40
  "scale_embedding": true,
41
+ "src_vocab_size": 42024,
42
+ "tgt_vocab_size": 42024,
43
  "tie_word_embeddings": true,
44
+ "torch_dtype": "float32",
45
+ "transformers_version": "4.53.0.dev0",
46
+ "use_cache": true
47
+ }
generation_config.json CHANGED
@@ -8,5 +8,5 @@
8
  "max_length": 200,
9
  "num_beams": 5,
10
  "pad_token_id": 1,
11
- "transformers_version": "4.27.0.dev0"
12
  }
 
8
  "max_length": 200,
9
  "num_beams": 5,
10
  "pad_token_id": 1,
11
+ "transformers_version": "4.53.0.dev0"
12
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:45fc9b0473baf52a55c8bcb2fcbf07e6d446b032fd66740380cf5676271c81e0
3
- size 1079014512
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0576d5db6ebd735176fc84920748d5812dd0e5ddda5c90998add52dc5b6c5a04
3
+ size 1087419504