Upload 6 files
Browse files- README.md +9 -3
- adapter_config.json +26 -0
- adapter_model.bin +3 -0
- config.json +74 -0
- non_lora_trainables.bin +3 -0
- trainer_state.json +0 -0
README.md
CHANGED
@@ -1,3 +1,9 @@
|
|
1 |
-
---
|
2 |
-
|
3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: peft
|
3 |
+
---
|
4 |
+
## Training procedure
|
5 |
+
|
6 |
+
### Framework versions
|
7 |
+
|
8 |
+
|
9 |
+
- PEFT 0.4.0
|
adapter_config.json
ADDED
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "liuhaotian/llava-v1.6-34b",
|
4 |
+
"bias": "none",
|
5 |
+
"fan_in_fan_out": false,
|
6 |
+
"inference_mode": true,
|
7 |
+
"init_lora_weights": true,
|
8 |
+
"layers_pattern": null,
|
9 |
+
"layers_to_transform": null,
|
10 |
+
"lora_alpha": 32,
|
11 |
+
"lora_dropout": 0.05,
|
12 |
+
"modules_to_save": null,
|
13 |
+
"peft_type": "LORA",
|
14 |
+
"r": 16,
|
15 |
+
"revision": null,
|
16 |
+
"target_modules": [
|
17 |
+
"gate_proj",
|
18 |
+
"q_proj",
|
19 |
+
"k_proj",
|
20 |
+
"up_proj",
|
21 |
+
"v_proj",
|
22 |
+
"down_proj",
|
23 |
+
"o_proj"
|
24 |
+
],
|
25 |
+
"task_type": "CAUSAL_LM"
|
26 |
+
}
|
adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9c361ee1f3f694130b56ad1cdde25a6fb94d7b4658574a49e53eb6c51ec2227d
|
3 |
+
size 250827242
|
config.json
ADDED
@@ -0,0 +1,74 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "liuhaotian/llava-v1.6-34b",
|
3 |
+
"architectures": [
|
4 |
+
"LlavaLlamaForCausalLM"
|
5 |
+
],
|
6 |
+
"attention_bias": false,
|
7 |
+
"attention_dropout": 0.0,
|
8 |
+
"bos_token_id": 1,
|
9 |
+
"eos_token_id": 7,
|
10 |
+
"freeze_mm_mlp_adapter": false,
|
11 |
+
"freeze_mm_vision_resampler": false,
|
12 |
+
"hidden_act": "silu",
|
13 |
+
"hidden_size": 7168,
|
14 |
+
"image_aspect_ratio": "anyres",
|
15 |
+
"image_crop_resolution": 224,
|
16 |
+
"image_grid_pinpoints": [
|
17 |
+
[
|
18 |
+
336,
|
19 |
+
672
|
20 |
+
],
|
21 |
+
[
|
22 |
+
672,
|
23 |
+
336
|
24 |
+
],
|
25 |
+
[
|
26 |
+
672,
|
27 |
+
672
|
28 |
+
],
|
29 |
+
[
|
30 |
+
1008,
|
31 |
+
336
|
32 |
+
],
|
33 |
+
[
|
34 |
+
336,
|
35 |
+
1008
|
36 |
+
]
|
37 |
+
],
|
38 |
+
"image_split_resolution": 224,
|
39 |
+
"image_token_index": 64002,
|
40 |
+
"initializer_range": 0.02,
|
41 |
+
"intermediate_size": 20480,
|
42 |
+
"max_position_embeddings": 4096,
|
43 |
+
"mm_hidden_size": 1024,
|
44 |
+
"mm_patch_merge_type": "spatial_unpad",
|
45 |
+
"mm_projector_lr": 2e-05,
|
46 |
+
"mm_projector_type": "mlp2x_gelu",
|
47 |
+
"mm_resampler_type": null,
|
48 |
+
"mm_use_im_patch_token": false,
|
49 |
+
"mm_use_im_start_end": false,
|
50 |
+
"mm_vision_select_feature": "patch",
|
51 |
+
"mm_vision_select_layer": -2,
|
52 |
+
"mm_vision_tower": "openai/clip-vit-large-patch14-336",
|
53 |
+
"mm_vision_tower_lr": 2e-06,
|
54 |
+
"model_type": "llava_llama",
|
55 |
+
"num_attention_heads": 56,
|
56 |
+
"num_hidden_layers": 60,
|
57 |
+
"num_key_value_heads": 8,
|
58 |
+
"pad_token_id": 0,
|
59 |
+
"pretraining_tp": 1,
|
60 |
+
"rms_norm_eps": 1e-05,
|
61 |
+
"rope_scaling": null,
|
62 |
+
"rope_theta": 5000000.0,
|
63 |
+
"tie_word_embeddings": false,
|
64 |
+
"tokenizer_model_max_length": 4096,
|
65 |
+
"tokenizer_padding_side": "right",
|
66 |
+
"torch_dtype": "bfloat16",
|
67 |
+
"transformers_version": "4.37.2",
|
68 |
+
"tune_mm_mlp_adapter": false,
|
69 |
+
"tune_mm_vision_resampler": false,
|
70 |
+
"unfreeze_mm_vision_tower": true,
|
71 |
+
"use_cache": true,
|
72 |
+
"use_mm_proj": true,
|
73 |
+
"vocab_size": 64000
|
74 |
+
}
|
non_lora_trainables.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ccfaeac2d55df47dbb4c461ff82dcf6bd9be26741d94b060c882082b78939db7
|
3 |
+
size 117471408
|
trainer_state.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|