Hugging Face's logo

taesiri
/
dummymodel2

PEFT
llava
4-bit precision
Model card Files Community
taesiri commited on
Commit
929b9a1
•
1 Parent(s): 59d4868

Upload 11 files

Browse files
Files changed (11) hide show
  1. README.md +20 -0
  2. adapter_config.json +26 -0
  3. adapter_model.bin +3 -0
  4. best_llava_eval_model/README.md +20 -0
  5. best_llava_eval_model/adapter_config.json +26 -0
  6. best_llava_eval_model/adapter_model.bin +3 -0
  7. best_llava_eval_model/config.json +58 -0
  8. best_llava_eval_model/non_lora_trainables.bin +3 -0
  9. config.json +58 -0
  10. non_lora_trainables.bin +3 -0
  11. trainer_state.json +0 -0
README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: ['mm_projector']
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: bfloat16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "./temp/llava-v1.5-13b",
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": 256,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 128,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "k_proj",
18
+ "v_proj",
19
+ "down_proj",
20
+ "q_proj",
21
+ "o_proj",
22
+ "gate_proj",
23
+ "up_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 sha256:111721fcfeefde2123f29b947f4e5c04aac315b02dcc2ee0ca10d2631a2e8da3
3
+ size 1001584333
best_llava_eval_model/README.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ ---
4
+ ## Training procedure
5
+
6
+
7
+ The following `bitsandbytes` quantization config was used during training:
8
+ - load_in_8bit: False
9
+ - load_in_4bit: True
10
+ - llm_int8_threshold: 6.0
11
+ - llm_int8_skip_modules: ['mm_projector']
12
+ - llm_int8_enable_fp32_cpu_offload: False
13
+ - llm_int8_has_fp16_weight: False
14
+ - bnb_4bit_quant_type: nf4
15
+ - bnb_4bit_use_double_quant: True
16
+ - bnb_4bit_compute_dtype: bfloat16
17
+ ### Framework versions
18
+
19
+
20
+ - PEFT 0.4.0
best_llava_eval_model/adapter_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_mapping": null,
3
+ "base_model_name_or_path": "./temp/llava-v1.5-13b",
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": 256,
11
+ "lora_dropout": 0.05,
12
+ "modules_to_save": null,
13
+ "peft_type": "LORA",
14
+ "r": 128,
15
+ "revision": null,
16
+ "target_modules": [
17
+ "k_proj",
18
+ "v_proj",
19
+ "down_proj",
20
+ "q_proj",
21
+ "o_proj",
22
+ "gate_proj",
23
+ "up_proj"
24
+ ],
25
+ "task_type": "CAUSAL_LM"
26
+ }
best_llava_eval_model/adapter_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:111721fcfeefde2123f29b947f4e5c04aac315b02dcc2ee0ca10d2631a2e8da3
3
+ size 1001584333
best_llava_eval_model/config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "./temp/llava-v1.5-13b",
3
+ "architectures": [
4
+ "LlavaLlamaForCausalLM"
5
+ ],
6
+ "bos_token_id": 1,
7
+ "eos_token_id": 2,
8
+ "freeze_mm_mlp_adapter": false,
9
+ "freeze_mm_vision_resampler": false,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 5120,
12
+ "image_aspect_ratio": "pad",
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 13824,
15
+ "max_length": 4096,
16
+ "max_position_embeddings": 4096,
17
+ "mm_hidden_size": 1024,
18
+ "mm_projector_lr": 2e-05,
19
+ "mm_projector_type": "mlp2x_gelu",
20
+ "mm_resampler_type": null,
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",
27
+ "num_attention_heads": 40,
28
+ "num_hidden_layers": 40,
29
+ "num_key_value_heads": 40,
30
+ "pad_token_id": 0,
31
+ "pretraining_tp": 1,
32
+ "quantization_config": {
33
+ "bnb_4bit_compute_dtype": "bfloat16",
34
+ "bnb_4bit_quant_type": "nf4",
35
+ "bnb_4bit_use_double_quant": true,
36
+ "llm_int8_enable_fp32_cpu_offload": false,
37
+ "llm_int8_has_fp16_weight": false,
38
+ "llm_int8_skip_modules": [
39
+ "mm_projector"
40
+ ],
41
+ "llm_int8_threshold": 6.0,
42
+ "load_in_4bit": true,
43
+ "load_in_8bit": false
44
+ },
45
+ "rms_norm_eps": 1e-05,
46
+ "rope_scaling": null,
47
+ "tie_word_embeddings": false,
48
+ "tokenizer_model_max_length": 2048,
49
+ "tokenizer_padding_side": "right",
50
+ "torch_dtype": "bfloat16",
51
+ "transformers_version": "4.31.0",
52
+ "tune_mm_mlp_adapter": false,
53
+ "tune_mm_vision_resampler": false,
54
+ "unfreeze_mm_vision_tower": false,
55
+ "use_cache": false,
56
+ "use_mm_proj": true,
57
+ "vocab_size": 32000
58
+ }
best_llava_eval_model/non_lora_trainables.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8527d5bb84baee7fcd5c1e82aa1220d6e7566c4b01b2d1601076be85c278c90a
3
+ size 62936807
config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "./temp/llava-v1.5-13b",
3
+ "architectures": [
4
+ "LlavaLlamaForCausalLM"
5
+ ],
6
+ "bos_token_id": 1,
7
+ "eos_token_id": 2,
8
+ "freeze_mm_mlp_adapter": false,
9
+ "freeze_mm_vision_resampler": false,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 5120,
12
+ "image_aspect_ratio": "pad",
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 13824,
15
+ "max_length": 4096,
16
+ "max_position_embeddings": 4096,
17
+ "mm_hidden_size": 1024,
18
+ "mm_projector_lr": 2e-05,
19
+ "mm_projector_type": "mlp2x_gelu",
20
+ "mm_resampler_type": null,
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",
27
+ "num_attention_heads": 40,
28
+ "num_hidden_layers": 40,
29
+ "num_key_value_heads": 40,
30
+ "pad_token_id": 0,
31
+ "pretraining_tp": 1,
32
+ "quantization_config": {
33
+ "bnb_4bit_compute_dtype": "bfloat16",
34
+ "bnb_4bit_quant_type": "nf4",
35
+ "bnb_4bit_use_double_quant": true,
36
+ "llm_int8_enable_fp32_cpu_offload": false,
37
+ "llm_int8_has_fp16_weight": false,
38
+ "llm_int8_skip_modules": [
39
+ "mm_projector"
40
+ ],
41
+ "llm_int8_threshold": 6.0,
42
+ "load_in_4bit": true,
43
+ "load_in_8bit": false
44
+ },
45
+ "rms_norm_eps": 1e-05,
46
+ "rope_scaling": null,
47
+ "tie_word_embeddings": false,
48
+ "tokenizer_model_max_length": 2048,
49
+ "tokenizer_padding_side": "right",
50
+ "torch_dtype": "bfloat16",
51
+ "transformers_version": "4.31.0",
52
+ "tune_mm_mlp_adapter": false,
53
+ "tune_mm_vision_resampler": false,
54
+ "unfreeze_mm_vision_tower": false,
55
+ "use_cache": true,
56
+ "use_mm_proj": true,
57
+ "vocab_size": 32000
58
+ }
non_lora_trainables.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8527d5bb84baee7fcd5c1e82aa1220d6e7566c4b01b2d1601076be85c278c90a
3
+ size 62936807
trainer_state.json ADDED
The diff for this file is too large to render. See raw diff