ahmedabdelwahed commited on
Commit
ae9b8ae
1 Parent(s): 31db12d

Upload LlavaMistralForCausalLM

Browse files
config.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/content/model-hf-repo",
3
+ "architectures": [
4
+ "LlavaMistralForCausalLM"
5
+ ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 1,
8
+ "eos_token_id": 2,
9
+ "feature_outs": "encoder+decoder",
10
+ "freeze_mm_mlp_adapter": false,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "image_aspect_ratio": "pad",
14
+ "img_size": 640,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 14336,
17
+ "max_position_embeddings": 32768,
18
+ "mm_hidden_size": 1024,
19
+ "mm_projector_lr": null,
20
+ "mm_projector_type": "mlp2x_gelu",
21
+ "mm_use_im_patch_token": false,
22
+ "mm_use_im_start_end": false,
23
+ "mm_vision_select_feature": "patch",
24
+ "mm_vision_select_layer": -2,
25
+ "mm_vision_tower": "openai/clip-vit-large-patch14-336",
26
+ "model_type": "llava_mistral",
27
+ "num_attention_heads": 32,
28
+ "num_hidden_layers": 32,
29
+ "num_key_value_heads": 8,
30
+ "proj_vis_to_txt_tokens": false,
31
+ "prompt_segtok_w_instruct": false,
32
+ "quantization_config": {
33
+ "_load_in_4bit": false,
34
+ "_load_in_8bit": true,
35
+ "bnb_4bit_compute_dtype": "float32",
36
+ "bnb_4bit_quant_storage": "uint8",
37
+ "bnb_4bit_quant_type": "fp4",
38
+ "bnb_4bit_use_double_quant": false,
39
+ "llm_int8_enable_fp32_cpu_offload": false,
40
+ "llm_int8_has_fp16_weight": false,
41
+ "llm_int8_skip_modules": null,
42
+ "llm_int8_threshold": 6.0,
43
+ "load_in_4bit": false,
44
+ "load_in_8bit": true,
45
+ "quant_method": "bitsandbytes"
46
+ },
47
+ "rms_norm_eps": 1e-05,
48
+ "rope_theta": 1000000.0,
49
+ "segtok_posembed": "sincos",
50
+ "sliding_window": null,
51
+ "tie_word_embeddings": false,
52
+ "tokenizer_model_max_length": 2048,
53
+ "tokenizer_padding_side": "right",
54
+ "torch_dtype": "float16",
55
+ "transformers_version": "4.42.4",
56
+ "tune_mm_mlp_adapter": false,
57
+ "tune_vision_tokenizer": "none",
58
+ "use_cache": true,
59
+ "use_mm_proj": true,
60
+ "vision_backbone": "convnextlarge",
61
+ "vision_tokenizer_lr": null,
62
+ "vocab_size": 32000
63
+ }
generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 2,
5
+ "transformers_version": "4.42.4"
6
+ }
model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:614487b9c1227c6deb74b6f3b68839f4a395fb9b2dfa12bdca2cbfbf6ca832d5
3
+ size 4947097024
model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d6e1b684636aa2ac932e1d1a3aefe884af4e55b160214d93ee13f99f09094e53
3
+ size 3797743050
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff