Hugging Face's logo

Nagase-Kotono
/
LLaVA-42dot_LLM-1.3B-pretrain

Text Generation
Transformers
Korean
English
llava_llama
Inference Endpoints
Model card Files Community
Nagase-Kotono commited on
Commit
a424b3c
•
1 Parent(s): 153ed66

03.22.15:27

Browse files
Files changed (9) hide show
  1. checkpoint-2000/config.json +48 -0
  2. checkpoint-2000/mm_projector.bin +3 -0
  3. checkpoint-4000/config.json +48 -0
  4. checkpoint-4000/mm_projector.bin +3 -0
  5. checkpoint-6000/config.json +48 -0
  6. checkpoint-6000/mm_projector.bin +3 -0
  7. config.json +48 -0
  8. mm_projector.bin +3 -0
  9. trainer_state.json +0 -0
checkpoint-2000/config.json ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "42dot/42dot_LLM-SFT-1.3B",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 50257,
9
+ "eos_token_id": 50256,
10
+ "freeze_mm_mlp_adapter": false,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "image_aspect_ratio": "square",
14
+ "image_projector_type": "mlp2x_gelu",
15
+ "initializer_range": 0.01,
16
+ "intermediate_size": 5632,
17
+ "max_position_embeddings": 4096,
18
+ "mm_hidden_size": 1024,
19
+ "mm_image_tower": "openai/clip-vit-large-patch14-336",
20
+ "mm_projector_lr": null,
21
+ "mm_use_im_patch_token": false,
22
+ "mm_use_im_start_end": false,
23
+ "mm_video_tower": null,
24
+ "mm_vision_select_feature": "patch",
25
+ "mm_vision_select_layer": -2,
26
+ "model_type": "llava_llama",
27
+ "num_attention_heads": 32,
28
+ "num_hidden_layers": 24,
29
+ "num_key_value_heads": 32,
30
+ "pad_token_id": 50259,
31
+ "pretraining_tp": 1,
32
+ "rms_norm_eps": 1e-06,
33
+ "rope_scaling": null,
34
+ "rope_theta": 10000.0,
35
+ "tie_word_embeddings": false,
36
+ "tokenizer_class": "GPT2TokenizerFast",
37
+ "tokenizer_padding_side": "right",
38
+ "torch_dtype": "float32",
39
+ "transformers_version": "4.37.0",
40
+ "tune_mm_mlp_adapter": true,
41
+ "use_cache": false,
42
+ "use_mm_proj": true,
43
+ "video_global_proj": false,
44
+ "video_projector_type": "linear",
45
+ "video_spatial_proj": false,
46
+ "video_temproal_proj": false,
47
+ "vocab_size": 50304
48
+ }
checkpoint-2000/mm_projector.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4f4b601c0fb2e65898a67a8a6b4feb93befcb66a3235075535d470e28018df17
3
+ size 12592829
checkpoint-4000/config.json ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "42dot/42dot_LLM-SFT-1.3B",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 50257,
9
+ "eos_token_id": 50256,
10
+ "freeze_mm_mlp_adapter": false,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "image_aspect_ratio": "square",
14
+ "image_projector_type": "mlp2x_gelu",
15
+ "initializer_range": 0.01,
16
+ "intermediate_size": 5632,
17
+ "max_position_embeddings": 4096,
18
+ "mm_hidden_size": 1024,
19
+ "mm_image_tower": "openai/clip-vit-large-patch14-336",
20
+ "mm_projector_lr": null,
21
+ "mm_use_im_patch_token": false,
22
+ "mm_use_im_start_end": false,
23
+ "mm_video_tower": null,
24
+ "mm_vision_select_feature": "patch",
25
+ "mm_vision_select_layer": -2,
26
+ "model_type": "llava_llama",
27
+ "num_attention_heads": 32,
28
+ "num_hidden_layers": 24,
29
+ "num_key_value_heads": 32,
30
+ "pad_token_id": 50259,
31
+ "pretraining_tp": 1,
32
+ "rms_norm_eps": 1e-06,
33
+ "rope_scaling": null,
34
+ "rope_theta": 10000.0,
35
+ "tie_word_embeddings": false,
36
+ "tokenizer_class": "GPT2TokenizerFast",
37
+ "tokenizer_padding_side": "right",
38
+ "torch_dtype": "float32",
39
+ "transformers_version": "4.37.0",
40
+ "tune_mm_mlp_adapter": true,
41
+ "use_cache": false,
42
+ "use_mm_proj": true,
43
+ "video_global_proj": false,
44
+ "video_projector_type": "linear",
45
+ "video_spatial_proj": false,
46
+ "video_temproal_proj": false,
47
+ "vocab_size": 50304
48
+ }
checkpoint-4000/mm_projector.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c289b245f583932e318a094fd6e02c0a7531344f70f1cbc235e6a77650a72f5
3
+ size 12592829
checkpoint-6000/config.json ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "42dot/42dot_LLM-SFT-1.3B",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 50257,
9
+ "eos_token_id": 50256,
10
+ "freeze_mm_mlp_adapter": false,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "image_aspect_ratio": "square",
14
+ "image_projector_type": "mlp2x_gelu",
15
+ "initializer_range": 0.01,
16
+ "intermediate_size": 5632,
17
+ "max_position_embeddings": 4096,
18
+ "mm_hidden_size": 1024,
19
+ "mm_image_tower": "openai/clip-vit-large-patch14-336",
20
+ "mm_projector_lr": null,
21
+ "mm_use_im_patch_token": false,
22
+ "mm_use_im_start_end": false,
23
+ "mm_video_tower": null,
24
+ "mm_vision_select_feature": "patch",
25
+ "mm_vision_select_layer": -2,
26
+ "model_type": "llava_llama",
27
+ "num_attention_heads": 32,
28
+ "num_hidden_layers": 24,
29
+ "num_key_value_heads": 32,
30
+ "pad_token_id": 50259,
31
+ "pretraining_tp": 1,
32
+ "rms_norm_eps": 1e-06,
33
+ "rope_scaling": null,
34
+ "rope_theta": 10000.0,
35
+ "tie_word_embeddings": false,
36
+ "tokenizer_class": "GPT2TokenizerFast",
37
+ "tokenizer_padding_side": "right",
38
+ "torch_dtype": "float32",
39
+ "transformers_version": "4.37.0",
40
+ "tune_mm_mlp_adapter": true,
41
+ "use_cache": false,
42
+ "use_mm_proj": true,
43
+ "video_global_proj": false,
44
+ "video_projector_type": "linear",
45
+ "video_spatial_proj": false,
46
+ "video_temproal_proj": false,
47
+ "vocab_size": 50304
48
+ }
checkpoint-6000/mm_projector.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0187d20fb39a446c7b11fadaff3e062196c27ed1ea48bfe257db4ec5a40d7c98
3
+ size 12592829
config.json ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "42dot/42dot_LLM-SFT-1.3B",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 50257,
9
+ "eos_token_id": 50256,
10
+ "freeze_mm_mlp_adapter": false,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "image_aspect_ratio": "square",
14
+ "image_projector_type": "mlp2x_gelu",
15
+ "initializer_range": 0.01,
16
+ "intermediate_size": 5632,
17
+ "max_position_embeddings": 4096,
18
+ "mm_hidden_size": 1024,
19
+ "mm_image_tower": "openai/clip-vit-large-patch14-336",
20
+ "mm_projector_lr": null,
21
+ "mm_use_im_patch_token": false,
22
+ "mm_use_im_start_end": false,
23
+ "mm_video_tower": null,
24
+ "mm_vision_select_feature": "patch",
25
+ "mm_vision_select_layer": -2,
26
+ "model_type": "llava_llama",
27
+ "num_attention_heads": 32,
28
+ "num_hidden_layers": 24,
29
+ "num_key_value_heads": 32,
30
+ "pad_token_id": 50259,
31
+ "pretraining_tp": 1,
32
+ "rms_norm_eps": 1e-06,
33
+ "rope_scaling": null,
34
+ "rope_theta": 10000.0,
35
+ "tie_word_embeddings": false,
36
+ "tokenizer_class": "GPT2TokenizerFast",
37
+ "tokenizer_padding_side": "right",
38
+ "torch_dtype": "float32",
39
+ "transformers_version": "4.37.0",
40
+ "tune_mm_mlp_adapter": true,
41
+ "use_cache": true,
42
+ "use_mm_proj": true,
43
+ "video_global_proj": false,
44
+ "video_projector_type": "linear",
45
+ "video_spatial_proj": false,
46
+ "video_temproal_proj": false,
47
+ "vocab_size": 50304
48
+ }
mm_projector.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a00d0f3aaecf58b5fb47a7f5dbb8b05667a1a0feafe3c60e4326ef6010e566e6
3
+ size 12592829
trainer_state.json ADDED
The diff for this file is too large to render. See raw diff