Andy Janco commited on
Commit
e49c310
1 Parent(s): c2b5b61

Upload Qwen2VLForConditionalGeneration

Browse files
Files changed (2) hide show
  1. config.json +9 -32
  2. model.safetensors +3 -0
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "Qwen/Qwen2-VL-7B-Instruct-GPTQ-Int4",
3
  "architectures": [
4
  "Qwen2VLForConditionalGeneration"
5
  ],
@@ -7,40 +7,16 @@
7
  "bos_token_id": 151643,
8
  "eos_token_id": 151645,
9
  "hidden_act": "silu",
10
- "hidden_size": 3584,
11
  "image_token_id": 151655,
12
  "initializer_range": 0.02,
13
- "intermediate_size": 18944,
14
  "max_position_embeddings": 32768,
15
  "max_window_layers": 28,
16
  "model_type": "qwen2_vl",
17
- "num_attention_heads": 28,
18
  "num_hidden_layers": 28,
19
- "num_key_value_heads": 4,
20
- "quantization_config": {
21
- "batch_size": 1,
22
- "bits": 4,
23
- "block_name_to_quantize": null,
24
- "cache_block_outputs": true,
25
- "damp_percent": 0.1,
26
- "dataset": null,
27
- "desc_act": false,
28
- "exllama_config": {
29
- "version": 1
30
- },
31
- "group_size": 128,
32
- "max_input_length": null,
33
- "model_seqlen": null,
34
- "module_name_preceding_first_block": null,
35
- "modules_in_block_to_quantize": null,
36
- "pad_token_id": null,
37
- "quant_method": "gptq",
38
- "sym": true,
39
- "tokenizer": null,
40
- "true_sequential": true,
41
- "use_cuda_fp16": false,
42
- "use_exllama": true
43
- },
44
  "rms_norm_eps": 1e-06,
45
  "rope_scaling": {
46
  "mrope_section": [
@@ -53,13 +29,14 @@
53
  },
54
  "rope_theta": 1000000.0,
55
  "sliding_window": 32768,
56
- "tie_word_embeddings": false,
57
- "torch_dtype": "float16",
58
  "transformers_version": "4.45.0.dev0",
59
  "use_cache": true,
60
  "use_sliding_window": false,
61
  "video_token_id": 151656,
62
  "vision_config": {
 
63
  "in_chans": 3,
64
  "model_type": "qwen2_vl",
65
  "spatial_patch_size": 14
@@ -67,5 +44,5 @@
67
  "vision_end_token_id": 151653,
68
  "vision_start_token_id": 151652,
69
  "vision_token_id": 151654,
70
- "vocab_size": 152064
71
  }
 
1
  {
2
+ "_name_or_path": "Qwen/Qwen2-VL-2B-Instruct",
3
  "architectures": [
4
  "Qwen2VLForConditionalGeneration"
5
  ],
 
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
  "rms_norm_eps": 1e-06,
21
  "rope_scaling": {
22
  "mrope_section": [
 
29
  },
30
  "rope_theta": 1000000.0,
31
  "sliding_window": 32768,
32
+ "tie_word_embeddings": true,
33
+ "torch_dtype": "bfloat16",
34
  "transformers_version": "4.45.0.dev0",
35
  "use_cache": true,
36
  "use_sliding_window": false,
37
  "video_token_id": 151656,
38
  "vision_config": {
39
+ "hidden_size": 1536,
40
  "in_chans": 3,
41
  "model_type": "qwen2_vl",
42
  "spatial_patch_size": 14
 
44
  "vision_end_token_id": 151653,
45
  "vision_start_token_id": 151652,
46
  "vision_token_id": 151654,
47
+ "vocab_size": 151936
48
  }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80c3c3e3afccee8ddb90a507aecc9949cfc3f8c51890d59049ef1f8f8683be5a
3
+ size 4418050848