new5558 commited on
Commit
fc2f55d
·
verified ·
1 Parent(s): 532919a

(Trained with Unsloth)

Browse files
Files changed (1) hide show
  1. config.json +48 -60
config.json CHANGED
@@ -1,64 +1,52 @@
1
  {
2
- "_name_or_path": "/content/unsloth_finetune",
3
- "architectures": [
4
- "Qwen2VLForConditionalGeneration"
5
- ],
6
- "attention_dropout": 0.0,
7
- "bos_token_id": 151643,
8
- "eos_token_id": 151645,
9
- "hidden_act": "silu",
10
- "hidden_size": 1536,
11
- "image_token_id": 151655,
12
- "initializer_range": 0.02,
13
- "intermediate_size": 8960,
14
- "max_position_embeddings": 32768,
15
- "max_window_layers": 28,
16
- "model_type": "qwen2_vl",
17
- "num_attention_heads": 12,
18
- "num_hidden_layers": 28,
19
- "num_key_value_heads": 2,
20
- "pad_token_id": 151654,
21
- "quantization_config": {
22
- "bnb_4bit_compute_dtype": "float16",
23
- "bnb_4bit_quant_type": "nf4",
24
- "bnb_4bit_use_double_quant": true,
25
- "llm_int8_enable_fp32_cpu_offload": false,
26
- "llm_int8_has_fp16_weight": false,
27
- "llm_int8_skip_modules": null,
28
- "llm_int8_threshold": 6.0,
29
- "load_in_4bit": true,
30
- "load_in_8bit": false,
31
- "quant_method": "bitsandbytes"
32
- },
33
- "rms_norm_eps": 1e-06,
34
- "rope_scaling": {
35
- "mrope_section": [
36
- 16,
37
- 24,
38
- 24
39
  ],
40
- "rope_type": "default",
41
- "type": "default"
42
- },
43
- "rope_theta": 1000000.0,
44
- "sliding_window": 32768,
45
- "tie_word_embeddings": true,
46
- "torch_dtype": "float16",
47
- "transformers_version": "4.48.3",
48
- "unsloth_fixed": true,
49
- "unsloth_version": "2025.2.15",
50
- "use_cache": true,
51
- "use_sliding_window": false,
52
- "video_token_id": 151656,
53
- "vision_config": {
54
  "hidden_size": 1536,
55
- "in_chans": 3,
 
 
 
 
56
  "model_type": "qwen2_vl",
57
- "spatial_patch_size": 14,
58
- "torch_dtype": "float16"
59
- },
60
- "vision_end_token_id": 151653,
61
- "vision_start_token_id": 151652,
62
- "vision_token_id": 151654,
63
- "vocab_size": 151936
64
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  {
2
+ "_name_or_path": "unsloth/qwen2-vl-2b-instruct-unsloth-bnb-4bit",
3
+ "architectures": [
4
+ "Qwen2VLForConditionalGeneration"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5
  ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 151643,
8
+ "eos_token_id": 151645,
9
+ "hidden_act": "silu",
 
 
 
 
 
 
 
 
 
 
10
  "hidden_size": 1536,
11
+ "image_token_id": 151655,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 8960,
14
+ "max_position_embeddings": 32768,
15
+ "max_window_layers": 28,
16
  "model_type": "qwen2_vl",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 28,
19
+ "num_key_value_heads": 2,
20
+ "pad_token_id": 151654,
21
+ "rms_norm_eps": 1e-06,
22
+ "rope_scaling": {
23
+ "mrope_section": [
24
+ 16,
25
+ 24,
26
+ 24
27
+ ],
28
+ "rope_type": "default",
29
+ "type": "default"
30
+ },
31
+ "rope_theta": 1000000.0,
32
+ "sliding_window": 32768,
33
+ "tie_word_embeddings": true,
34
+ "torch_dtype": "float16",
35
+ "transformers_version": "4.48.3",
36
+ "unsloth_fixed": true,
37
+ "unsloth_version": "2025.2.15",
38
+ "use_cache": true,
39
+ "use_sliding_window": false,
40
+ "video_token_id": 151656,
41
+ "vision_config": {
42
+ "hidden_size": 1536,
43
+ "in_chans": 3,
44
+ "model_type": "qwen2_vl",
45
+ "spatial_patch_size": 14,
46
+ "torch_dtype": "float16"
47
+ },
48
+ "vision_end_token_id": 151653,
49
+ "vision_start_token_id": 151652,
50
+ "vision_token_id": 151654,
51
+ "vocab_size": 151936
52
+ }