RichardErkhov commited on
Commit
953ccc6
1 Parent(s): c426d0c

uploaded model

Browse files
Files changed (1) hide show
  1. config.json +75 -0
config.json ADDED
@@ -0,0 +1,75 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "mbart-large-en-ro",
3
+ "_num_labels": 3,
4
+ "activation_dropout": 0.0,
5
+ "activation_function": "gelu",
6
+ "add_bias_logits": false,
7
+ "add_final_layer_norm": true,
8
+ "architectures": [
9
+ "MBartForCausalLM"
10
+ ],
11
+ "attention_dropout": 0.1,
12
+ "bos_token_id": 0,
13
+ "classif_dropout": 0.1,
14
+ "classifier_dropout": 0.0,
15
+ "d_model": 1024,
16
+ "decoder_attention_heads": 16,
17
+ "decoder_ffn_dim": 4096,
18
+ "decoder_layerdrop": 0.0,
19
+ "decoder_layers": 12,
20
+ "decoder_start_token_id": 250020,
21
+ "do_blenderbot_90_layernorm": false,
22
+ "dropout": 0.1,
23
+ "encoder_attention_heads": 16,
24
+ "encoder_ffn_dim": 4096,
25
+ "encoder_layerdrop": 0.0,
26
+ "encoder_layers": 12,
27
+ "eos_token_id": 2,
28
+ "extra_pos_embeddings": 2,
29
+ "force_bos_token_to_be_generated": false,
30
+ "forced_eos_token_id": 2,
31
+ "gradient_checkpointing": false,
32
+ "id2label": {
33
+ "0": "LABEL_0",
34
+ "1": "LABEL_1",
35
+ "2": "LABEL_2"
36
+ },
37
+ "init_std": 0.02,
38
+ "is_decoder": true,
39
+ "is_encoder_decoder": false,
40
+ "label2id": {
41
+ "LABEL_0": 0,
42
+ "LABEL_1": 1,
43
+ "LABEL_2": 2
44
+ },
45
+ "max_length": 1024,
46
+ "max_position_embeddings": 1024,
47
+ "model_type": "mbart",
48
+ "normalize_before": true,
49
+ "normalize_embedding": true,
50
+ "num_beams": 5,
51
+ "num_hidden_layers": 12,
52
+ "output_past": true,
53
+ "pad_token_id": 1,
54
+ "quantization_config": {
55
+ "_load_in_4bit": true,
56
+ "_load_in_8bit": false,
57
+ "bnb_4bit_compute_dtype": "float32",
58
+ "bnb_4bit_quant_storage": "uint8",
59
+ "bnb_4bit_quant_type": "fp4",
60
+ "bnb_4bit_use_double_quant": false,
61
+ "llm_int8_enable_fp32_cpu_offload": false,
62
+ "llm_int8_has_fp16_weight": false,
63
+ "llm_int8_skip_modules": null,
64
+ "llm_int8_threshold": 6.0,
65
+ "load_in_4bit": true,
66
+ "load_in_8bit": false,
67
+ "quant_method": "bitsandbytes"
68
+ },
69
+ "scale_embedding": true,
70
+ "static_position_embeddings": false,
71
+ "torch_dtype": "float16",
72
+ "transformers_version": "4.40.2",
73
+ "use_cache": true,
74
+ "vocab_size": 250027
75
+ }