Hugging Face's logo

khazic
/
kkk

Model card Files
xet
Community
khazic commited on
Commit
3be578e
·
verified ·
1 Parent(s): 4efda66

Upload folder using huggingface_hub

Browse files
Files changed (5) hide show
  1. checkpoints-pretrain-bunny/bunny-llama3-8b-pretrain/checkpoint-3906/config.json +38 -0
  2. checkpoints-pretrain-bunny/bunny-llama3-8b-pretrain/checkpoint-3906/mm_projector.bin +3 -0
  3. checkpoints-pretrain-bunny/bunny-llama3-8b-pretrain/config.json +38 -0
  4. checkpoints-pretrain-bunny/bunny-llama3-8b-pretrain/mm_projector.bin +3 -0
  5. checkpoints-pretrain-bunny/bunny-llama3-8b-pretrain/trainer_state.json +0 -0
checkpoints-pretrain-bunny/bunny-llama3-8b-pretrain/checkpoint-3906/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/llm-align/liuchonghan/xiaomin/model/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": 1280,
18
+ "mm_projector_lr": null,
19
+ "mm_projector_type": "mlp2x_gelu",
20
+ "model_type": "bunny-llama",
21
+ "num_attention_heads": 32,
22
+ "num_hidden_layers": 32,
23
+ "num_key_value_heads": 8,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-05,
26
+ "rope_scaling": null,
27
+ "rope_theta": 500000.0,
28
+ "tie_word_embeddings": false,
29
+ "tokenizer_model_max_length": 2048,
30
+ "tokenizer_padding_side": "right",
31
+ "torch_dtype": "bfloat16",
32
+ "transformers_version": "4.44.0",
33
+ "tune_mm_mlp_adapter": true,
34
+ "use_cache": false,
35
+ "use_mm_proj": true,
36
+ "use_s2": false,
37
+ "vocab_size": 128256
38
+ }
checkpoints-pretrain-bunny/bunny-llama3-8b-pretrain/checkpoint-3906/mm_projector.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:568c8d86b4d10eabcaeec57dc1dd550e3cf56a31b944185b52973cfe9cd17b41
3
+ size 44058680
checkpoints-pretrain-bunny/bunny-llama3-8b-pretrain/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "/llm-align/liuchonghan/xiaomin/model/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": 1280,
18
+ "mm_projector_lr": null,
19
+ "mm_projector_type": "mlp2x_gelu",
20
+ "model_type": "bunny-llama",
21
+ "num_attention_heads": 32,
22
+ "num_hidden_layers": 32,
23
+ "num_key_value_heads": 8,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-05,
26
+ "rope_scaling": null,
27
+ "rope_theta": 500000.0,
28
+ "tie_word_embeddings": false,
29
+ "tokenizer_model_max_length": 2048,
30
+ "tokenizer_padding_side": "right",
31
+ "torch_dtype": "bfloat16",
32
+ "transformers_version": "4.44.0",
33
+ "tune_mm_mlp_adapter": true,
34
+ "use_cache": true,
35
+ "use_mm_proj": true,
36
+ "use_s2": false,
37
+ "vocab_size": 128256
38
+ }
checkpoints-pretrain-bunny/bunny-llama3-8b-pretrain/mm_projector.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:568c8d86b4d10eabcaeec57dc1dd550e3cf56a31b944185b52973cfe9cd17b41
3
+ size 44058680
checkpoints-pretrain-bunny/bunny-llama3-8b-pretrain/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff