mmaaz60 commited on
Commit
997f542
1 Parent(s): 5728d06

Upload folder using huggingface_hub

Browse files
Files changed (3) hide show
  1. config.json +44 -0
  2. mm_projector.bin +3 -0
  3. trainer_state.json +0 -0
config.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "pretrained_checkpoints/Meta-Llama-3-8B-Instruct",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 128000,
9
+ "eos_token_id": 128001,
10
+ "freeze_mm_mlp_adapter": false,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "image_aspect_ratio": "square",
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 14336,
16
+ "max_position_embeddings": 8192,
17
+ "mm_hidden_size": 1024,
18
+ "mm_patch_merge_type": "flat",
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_llama",
27
+ "num_attention_heads": 32,
28
+ "num_hidden_layers": 32,
29
+ "num_key_value_heads": 8,
30
+ "pad_token_id": 128256,
31
+ "pretraining_tp": 1,
32
+ "rms_norm_eps": 1e-05,
33
+ "rope_scaling": null,
34
+ "rope_theta": 500000.0,
35
+ "tie_word_embeddings": false,
36
+ "tokenizer_model_max_length": 2048,
37
+ "tokenizer_padding_side": "right",
38
+ "torch_dtype": "bfloat16",
39
+ "transformers_version": "4.41.0.dev0",
40
+ "tune_mm_mlp_adapter": true,
41
+ "use_cache": true,
42
+ "use_mm_proj": true,
43
+ "vocab_size": 128257
44
+ }
mm_projector.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:070667a843c6aeb9000f624bc1ce5284e63cb4220eaa0ab6437b2922802981f7
3
+ size 41961528
trainer_state.json ADDED
The diff for this file is too large to render. See raw diff