feat: update config.json
Browse files- .gitattributes +1 -11
- config.json +107 -3
.gitattributes
CHANGED
|
@@ -33,25 +33,15 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
-
added_tokens.json filter=lfs diff=lfs merge=lfs -text
|
| 37 |
merges.txt filter=lfs diff=lfs merge=lfs -text
|
| 38 |
model-00001-of-00004.safetensors filter=lfs diff=lfs merge=lfs -text
|
| 39 |
modeling_valley.py filter=lfs diff=lfs merge=lfs -text
|
| 40 |
-
chat_template.json filter=lfs diff=lfs merge=lfs -text
|
| 41 |
-
config.json filter=lfs diff=lfs merge=lfs -text
|
| 42 |
-
generation_config.json filter=lfs diff=lfs merge=lfs -text
|
| 43 |
modeling_projector.py filter=lfs diff=lfs merge=lfs -text
|
| 44 |
-
model.safetensors.index.json filter=lfs diff=lfs merge=lfs -text
|
| 45 |
-
preprocessor_config.json filter=lfs diff=lfs merge=lfs -text
|
| 46 |
processing_valley.py filter=lfs diff=lfs merge=lfs -text
|
| 47 |
-
README.md filter=lfs diff=lfs merge=lfs -text
|
| 48 |
-
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 49 |
model-00004-of-00004.safetensors filter=lfs diff=lfs merge=lfs -text
|
| 50 |
-
special_tokens_map.json filter=lfs diff=lfs merge=lfs -text
|
| 51 |
-
tokenizer_config.json filter=lfs diff=lfs merge=lfs -text
|
| 52 |
utils.py filter=lfs diff=lfs merge=lfs -text
|
| 53 |
-
vocab.json filter=lfs diff=lfs merge=lfs -text
|
| 54 |
chat_template.jinja filter=lfs diff=lfs merge=lfs -text
|
| 55 |
model-00002-of-00004.safetensors filter=lfs diff=lfs merge=lfs -text
|
| 56 |
model-00003-of-00004.safetensors filter=lfs diff=lfs merge=lfs -text
|
| 57 |
modeling_vision_tower.py filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
| 36 |
merges.txt filter=lfs diff=lfs merge=lfs -text
|
| 37 |
model-00001-of-00004.safetensors filter=lfs diff=lfs merge=lfs -text
|
| 38 |
modeling_valley.py filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
| 39 |
modeling_projector.py filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
| 40 |
processing_valley.py filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
| 41 |
model-00004-of-00004.safetensors filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
| 42 |
utils.py filter=lfs diff=lfs merge=lfs -text
|
|
|
|
| 43 |
chat_template.jinja filter=lfs diff=lfs merge=lfs -text
|
| 44 |
model-00002-of-00004.safetensors filter=lfs diff=lfs merge=lfs -text
|
| 45 |
model-00003-of-00004.safetensors filter=lfs diff=lfs merge=lfs -text
|
| 46 |
modeling_vision_tower.py filter=lfs diff=lfs merge=lfs -text
|
| 47 |
+
valley_structure.png filter=lfs diff=lfs merge=lfs -text
|
config.json
CHANGED
|
@@ -1,3 +1,107 @@
|
|
| 1 |
-
|
| 2 |
-
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"anyres": false,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"ValleyQwen3ForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"auto_map": {
|
| 9 |
+
"AutoConfig": "modeling_valley.ValleyConfig",
|
| 10 |
+
"AutoModel": "modeling_valley.ValleyQwen3ForCausalLM",
|
| 11 |
+
"AutoModelForCausalLM": "modeling_valley.ValleyQwen3ForCausalLM"
|
| 12 |
+
},
|
| 13 |
+
"dtype": "bfloat16",
|
| 14 |
+
"eagle_vision_tower": "Qwen/Qwen2-VL-7B-Instruct",
|
| 15 |
+
"enable_thinking": false,
|
| 16 |
+
"eos_token_id": 151645,
|
| 17 |
+
"frame_max_pixels": 153664,
|
| 18 |
+
"grid_pinpoints": "(1x1),...,(3x3)",
|
| 19 |
+
"head_dim": 128,
|
| 20 |
+
"hidden_act": "silu",
|
| 21 |
+
"hidden_size": 4096,
|
| 22 |
+
"image_aspect_ratio": null,
|
| 23 |
+
"image_grid_pinpoints": null,
|
| 24 |
+
"initializer_range": 0.02,
|
| 25 |
+
"intermediate_size": 12288,
|
| 26 |
+
"language": "english",
|
| 27 |
+
"layer_types": [
|
| 28 |
+
"full_attention",
|
| 29 |
+
"full_attention",
|
| 30 |
+
"full_attention",
|
| 31 |
+
"full_attention",
|
| 32 |
+
"full_attention",
|
| 33 |
+
"full_attention",
|
| 34 |
+
"full_attention",
|
| 35 |
+
"full_attention",
|
| 36 |
+
"full_attention",
|
| 37 |
+
"full_attention",
|
| 38 |
+
"full_attention",
|
| 39 |
+
"full_attention",
|
| 40 |
+
"full_attention",
|
| 41 |
+
"full_attention",
|
| 42 |
+
"full_attention",
|
| 43 |
+
"full_attention",
|
| 44 |
+
"full_attention",
|
| 45 |
+
"full_attention",
|
| 46 |
+
"full_attention",
|
| 47 |
+
"full_attention",
|
| 48 |
+
"full_attention",
|
| 49 |
+
"full_attention",
|
| 50 |
+
"full_attention",
|
| 51 |
+
"full_attention",
|
| 52 |
+
"full_attention",
|
| 53 |
+
"full_attention",
|
| 54 |
+
"full_attention",
|
| 55 |
+
"full_attention",
|
| 56 |
+
"full_attention",
|
| 57 |
+
"full_attention",
|
| 58 |
+
"full_attention",
|
| 59 |
+
"full_attention",
|
| 60 |
+
"full_attention",
|
| 61 |
+
"full_attention",
|
| 62 |
+
"full_attention",
|
| 63 |
+
"full_attention"
|
| 64 |
+
],
|
| 65 |
+
"llm_name": "qwen3",
|
| 66 |
+
"max_pixels": 1536640,
|
| 67 |
+
"max_position_embeddings": 32768,
|
| 68 |
+
"max_tile_num": 9,
|
| 69 |
+
"max_vision_token": 100000,
|
| 70 |
+
"max_window_layers": 36,
|
| 71 |
+
"min_tile_num": 1,
|
| 72 |
+
"mlp_hidden_dim": null,
|
| 73 |
+
"mm_hidden_size": null,
|
| 74 |
+
"mm_projector_type": "ovis_conv_adapter_navit",
|
| 75 |
+
"mm_use_im_patch_token": false,
|
| 76 |
+
"mm_use_im_start_end": false,
|
| 77 |
+
"mm_vision_select_feature": "cls_patch",
|
| 78 |
+
"mm_vision_select_layer": -2,
|
| 79 |
+
"mm_vision_tower": "navit",
|
| 80 |
+
"model_class": "valley-product",
|
| 81 |
+
"model_type": "valley",
|
| 82 |
+
"navit_merger_hidden_dim": 65536,
|
| 83 |
+
"navit_use_mm_projector": false,
|
| 84 |
+
"num_attention_heads": 32,
|
| 85 |
+
"num_hidden_layers": 36,
|
| 86 |
+
"num_key_value_heads": 8,
|
| 87 |
+
"only_crop_single_image": true,
|
| 88 |
+
"only_navit": true,
|
| 89 |
+
"pad_token_id": 151643,
|
| 90 |
+
"pixelshuffle_downsample_ratio": 3,
|
| 91 |
+
"pool_out_size": 8,
|
| 92 |
+
"pooling_stride": 1,
|
| 93 |
+
"rms_norm_eps": 1e-06,
|
| 94 |
+
"rope_scaling": null,
|
| 95 |
+
"rope_theta": 1000000,
|
| 96 |
+
"sliding_window": null,
|
| 97 |
+
"tie_word_embeddings": false,
|
| 98 |
+
"tokenize_function": "softmax",
|
| 99 |
+
"torch_dtype": "bfloat16",
|
| 100 |
+
"training_stage": "stage2",
|
| 101 |
+
"transformers_version": "4.54.0",
|
| 102 |
+
"tune_mm_mlp_adapter": false,
|
| 103 |
+
"use_cache": false,
|
| 104 |
+
"use_sliding_window": false,
|
| 105 |
+
"use_special_start_end_token": true,
|
| 106 |
+
"vocab_size": 151675
|
| 107 |
+
}
|