Mediocreatmybest commited on
Commit
cb24198
1 Parent(s): 6613084

Upload InstructBlipForConditionalGeneration

Browse files
Files changed (2) hide show
  1. config.json +97 -0
  2. pytorch_model.bin +3 -0
config.json ADDED
@@ -0,0 +1,97 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Salesforce/instructblip-flan-t5-xl",
3
+ "architectures": [
4
+ "InstructBlipForConditionalGeneration"
5
+ ],
6
+ "initializer_factor": 1.0,
7
+ "initializer_range": 0.02,
8
+ "is_encoder_decoder": true,
9
+ "model_type": "instructblip",
10
+ "num_query_tokens": 32,
11
+ "qformer_config": {
12
+ "classifier_dropout": null,
13
+ "model_type": "instructblip_qformer",
14
+ "vocab_size": 30523
15
+ },
16
+ "quantization_config": {
17
+ "bnb_4bit_compute_dtype": "float32",
18
+ "bnb_4bit_quant_type": "fp4",
19
+ "bnb_4bit_use_double_quant": false,
20
+ "llm_int8_enable_fp32_cpu_offload": true,
21
+ "llm_int8_has_fp16_weight": false,
22
+ "llm_int8_skip_modules": null,
23
+ "llm_int8_threshold": 6.0,
24
+ "load_in_4bit": false,
25
+ "load_in_8bit": true
26
+ },
27
+ "text_config": {
28
+ "architectures": [
29
+ "T5ForConditionalGeneration"
30
+ ],
31
+ "bos_token_id": 1,
32
+ "classifier_dropout": 0.0,
33
+ "d_ff": 5120,
34
+ "d_kv": 64,
35
+ "d_model": 2048,
36
+ "decoder_start_token_id": 0,
37
+ "dense_act_fn": "gelu",
38
+ "dropout_rate": 0.1,
39
+ "eos_token_id": 1,
40
+ "feed_forward_proj": "gated-gelu",
41
+ "initializer_factor": 1.0,
42
+ "is_encoder_decoder": true,
43
+ "is_gated_act": true,
44
+ "layer_norm_epsilon": 1e-06,
45
+ "model_type": "t5",
46
+ "n_positions": 512,
47
+ "num_decoder_layers": 24,
48
+ "num_heads": 32,
49
+ "num_layers": 24,
50
+ "output_past": true,
51
+ "pad_token_id": 0,
52
+ "relative_attention_max_distance": 128,
53
+ "relative_attention_num_buckets": 32,
54
+ "task_specific_params": {
55
+ "summarization": {
56
+ "early_stopping": true,
57
+ "length_penalty": 2.0,
58
+ "max_length": 200,
59
+ "min_length": 30,
60
+ "no_repeat_ngram_size": 3,
61
+ "num_beams": 4,
62
+ "prefix": "summarize: "
63
+ },
64
+ "translation_en_to_de": {
65
+ "early_stopping": true,
66
+ "max_length": 300,
67
+ "num_beams": 4,
68
+ "prefix": "translate English to German: "
69
+ },
70
+ "translation_en_to_fr": {
71
+ "early_stopping": true,
72
+ "max_length": 300,
73
+ "num_beams": 4,
74
+ "prefix": "translate English to French: "
75
+ },
76
+ "translation_en_to_ro": {
77
+ "early_stopping": true,
78
+ "max_length": 300,
79
+ "num_beams": 4,
80
+ "prefix": "translate English to Romanian: "
81
+ }
82
+ },
83
+ "tie_word_embeddings": false,
84
+ "torch_dtype": "float32",
85
+ "vocab_size": 32128
86
+ },
87
+ "tie_word_embeddings": false,
88
+ "torch_dtype": "float16",
89
+ "transformers_version": "4.32.0.dev0",
90
+ "use_decoder_only_language_model": false,
91
+ "vision_config": {
92
+ "dropout": 0.0,
93
+ "initializer_factor": 1.0,
94
+ "model_type": "instructblip_vision_model",
95
+ "num_channels": 3
96
+ }
97
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:946a27f67b563504dd1a08b06940b32f6f79f0bc5be11567235cac1b38fd0dd3
3
+ size 5745359505