fix(text_encoder): restore Lightricks config with layer_types and full architecture fields
Browse files- text_encoder/config.json +76 -4
text_encoder/config.json
CHANGED
|
@@ -12,29 +12,101 @@
|
|
| 12 |
"initializer_range": 0.02,
|
| 13 |
"mm_tokens_per_image": 256,
|
| 14 |
"model_type": "gemma3",
|
|
|
|
| 15 |
"text_config": {
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
"hidden_size": 3840,
|
|
|
|
| 17 |
"intermediate_size": 15360,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
"model_type": "gemma3_text",
|
| 19 |
"num_attention_heads": 16,
|
| 20 |
"num_hidden_layers": 48,
|
| 21 |
"num_key_value_heads": 8,
|
|
|
|
|
|
|
|
|
|
| 22 |
"rope_scaling": {
|
| 23 |
"factor": 8.0,
|
| 24 |
"rope_type": "linear"
|
| 25 |
},
|
| 26 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
},
|
| 28 |
-
"
|
| 29 |
-
"transformers_version": "4.50.0.dev0",
|
| 30 |
"vision_config": {
|
|
|
|
|
|
|
| 31 |
"hidden_size": 1152,
|
| 32 |
"image_size": 896,
|
| 33 |
"intermediate_size": 4304,
|
|
|
|
| 34 |
"model_type": "siglip_vision_model",
|
| 35 |
"num_attention_heads": 16,
|
|
|
|
| 36 |
"num_hidden_layers": 27,
|
| 37 |
"patch_size": 14,
|
| 38 |
"vision_use_head": false
|
| 39 |
}
|
| 40 |
-
}
|
|
|
|
| 12 |
"initializer_range": 0.02,
|
| 13 |
"mm_tokens_per_image": 256,
|
| 14 |
"model_type": "gemma3",
|
| 15 |
+
"torch_dtype": "bfloat16",
|
| 16 |
"text_config": {
|
| 17 |
+
"_sliding_window_pattern": 6,
|
| 18 |
+
"attention_bias": false,
|
| 19 |
+
"attention_dropout": 0.0,
|
| 20 |
+
"attn_logit_softcapping": null,
|
| 21 |
+
"cache_implementation": "hybrid",
|
| 22 |
+
"final_logit_softcapping": null,
|
| 23 |
+
"head_dim": 256,
|
| 24 |
+
"hidden_activation": "gelu_pytorch_tanh",
|
| 25 |
"hidden_size": 3840,
|
| 26 |
+
"initializer_range": 0.02,
|
| 27 |
"intermediate_size": 15360,
|
| 28 |
+
"layer_types": [
|
| 29 |
+
"sliding_attention",
|
| 30 |
+
"sliding_attention",
|
| 31 |
+
"sliding_attention",
|
| 32 |
+
"sliding_attention",
|
| 33 |
+
"sliding_attention",
|
| 34 |
+
"full_attention",
|
| 35 |
+
"sliding_attention",
|
| 36 |
+
"sliding_attention",
|
| 37 |
+
"sliding_attention",
|
| 38 |
+
"sliding_attention",
|
| 39 |
+
"sliding_attention",
|
| 40 |
+
"full_attention",
|
| 41 |
+
"sliding_attention",
|
| 42 |
+
"sliding_attention",
|
| 43 |
+
"sliding_attention",
|
| 44 |
+
"sliding_attention",
|
| 45 |
+
"sliding_attention",
|
| 46 |
+
"full_attention",
|
| 47 |
+
"sliding_attention",
|
| 48 |
+
"sliding_attention",
|
| 49 |
+
"sliding_attention",
|
| 50 |
+
"sliding_attention",
|
| 51 |
+
"sliding_attention",
|
| 52 |
+
"full_attention",
|
| 53 |
+
"sliding_attention",
|
| 54 |
+
"sliding_attention",
|
| 55 |
+
"sliding_attention",
|
| 56 |
+
"sliding_attention",
|
| 57 |
+
"sliding_attention",
|
| 58 |
+
"full_attention",
|
| 59 |
+
"sliding_attention",
|
| 60 |
+
"sliding_attention",
|
| 61 |
+
"sliding_attention",
|
| 62 |
+
"sliding_attention",
|
| 63 |
+
"sliding_attention",
|
| 64 |
+
"full_attention",
|
| 65 |
+
"sliding_attention",
|
| 66 |
+
"sliding_attention",
|
| 67 |
+
"sliding_attention",
|
| 68 |
+
"sliding_attention",
|
| 69 |
+
"sliding_attention",
|
| 70 |
+
"full_attention",
|
| 71 |
+
"sliding_attention",
|
| 72 |
+
"sliding_attention",
|
| 73 |
+
"sliding_attention",
|
| 74 |
+
"sliding_attention",
|
| 75 |
+
"sliding_attention",
|
| 76 |
+
"full_attention"
|
| 77 |
+
],
|
| 78 |
+
"max_position_embeddings": 131072,
|
| 79 |
"model_type": "gemma3_text",
|
| 80 |
"num_attention_heads": 16,
|
| 81 |
"num_hidden_layers": 48,
|
| 82 |
"num_key_value_heads": 8,
|
| 83 |
+
"query_pre_attn_scalar": 256,
|
| 84 |
+
"rms_norm_eps": 1e-06,
|
| 85 |
+
"rope_local_base_freq": 10000,
|
| 86 |
"rope_scaling": {
|
| 87 |
"factor": 8.0,
|
| 88 |
"rope_type": "linear"
|
| 89 |
},
|
| 90 |
+
"rope_theta": 1000000,
|
| 91 |
+
"sliding_window": 1024,
|
| 92 |
+
"sliding_window_pattern": 6,
|
| 93 |
+
"use_bidirectional_attention": false,
|
| 94 |
+
"use_cache": true,
|
| 95 |
+
"vocab_size": 262208
|
| 96 |
},
|
| 97 |
+
"transformers_version": "4.57.3",
|
|
|
|
| 98 |
"vision_config": {
|
| 99 |
+
"attention_dropout": 0.0,
|
| 100 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 101 |
"hidden_size": 1152,
|
| 102 |
"image_size": 896,
|
| 103 |
"intermediate_size": 4304,
|
| 104 |
+
"layer_norm_eps": 1e-06,
|
| 105 |
"model_type": "siglip_vision_model",
|
| 106 |
"num_attention_heads": 16,
|
| 107 |
+
"num_channels": 3,
|
| 108 |
"num_hidden_layers": 27,
|
| 109 |
"patch_size": 14,
|
| 110 |
"vision_use_head": false
|
| 111 |
}
|
| 112 |
+
}
|