Update tiny models for NeoMMEForMaskedLM
Browse files- config.json +53 -0
- model.safetensors +3 -0
- preprocessor_config.json +20 -0
- tokenizer.json +0 -0
- tokenizer_config.json +28 -0
config.json
ADDED
|
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"allow_global_per_layer_attribute_access": true,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"NeoMMEForMaskedLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"document_token_id": 5,
|
| 9 |
+
"dtype": "float32",
|
| 10 |
+
"embedding_dim": 8,
|
| 11 |
+
"embedding_rank": 16,
|
| 12 |
+
"head_dim": 16,
|
| 13 |
+
"hidden_act": "relu2",
|
| 14 |
+
"hidden_size": 32,
|
| 15 |
+
"image_token_id": 6,
|
| 16 |
+
"initializer_range": 0.02,
|
| 17 |
+
"intermediate_size": 48,
|
| 18 |
+
"layer_types": [
|
| 19 |
+
"sliding_attention",
|
| 20 |
+
"full_attention"
|
| 21 |
+
],
|
| 22 |
+
"max_position_embeddings": 200,
|
| 23 |
+
"mlp_bias": false,
|
| 24 |
+
"model_type": "neomme",
|
| 25 |
+
"norm_eps": 1e-06,
|
| 26 |
+
"num_attention_heads": 4,
|
| 27 |
+
"num_hidden_layers": 2,
|
| 28 |
+
"num_key_value_heads": 2,
|
| 29 |
+
"pad_token_id": 0,
|
| 30 |
+
"patch_size": 4,
|
| 31 |
+
"per_layer_config": {
|
| 32 |
+
"1": {
|
| 33 |
+
"sliding_window": null
|
| 34 |
+
}
|
| 35 |
+
},
|
| 36 |
+
"residual_multiplier": 0.5,
|
| 37 |
+
"rope_parameters": {
|
| 38 |
+
"full_attention": {
|
| 39 |
+
"partial_rotary_factor": 0.25,
|
| 40 |
+
"rope_theta": 1000000.0,
|
| 41 |
+
"rope_type": "default"
|
| 42 |
+
},
|
| 43 |
+
"sliding_attention": {
|
| 44 |
+
"partial_rotary_factor": 1.0,
|
| 45 |
+
"rope_theta": 10000.0,
|
| 46 |
+
"rope_type": "default"
|
| 47 |
+
}
|
| 48 |
+
},
|
| 49 |
+
"sliding_window": 3,
|
| 50 |
+
"tie_word_embeddings": true,
|
| 51 |
+
"transformers_version": "5.19.0.dev0",
|
| 52 |
+
"vocab_size": 131072
|
| 53 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:defefe5623058c5f65cd2b755cbb81bce4a9ab29c19330baaf6999ddc2c3e313
|
| 3 |
+
size 25281848
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_rgb": true,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": true,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.5,
|
| 8 |
+
0.5,
|
| 9 |
+
0.5
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "NeoMMEImageProcessor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.5,
|
| 14 |
+
0.5,
|
| 15 |
+
0.5
|
| 16 |
+
],
|
| 17 |
+
"patch_size": 32,
|
| 18 |
+
"resample": 2,
|
| 19 |
+
"rescale_factor": 0.00392156862745098
|
| 20 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"clean_up_tokenization_spaces": false,
|
| 4 |
+
"document_token": "<doc>",
|
| 5 |
+
"eos_token": "<eos>",
|
| 6 |
+
"extra_special_tokens": [],
|
| 7 |
+
"image_token": "<img>",
|
| 8 |
+
"is_local": true,
|
| 9 |
+
"local_files_only": false,
|
| 10 |
+
"mask_token": "<mask>",
|
| 11 |
+
"model_input_names": [
|
| 12 |
+
"input_ids",
|
| 13 |
+
"attention_mask"
|
| 14 |
+
],
|
| 15 |
+
"model_max_length": 200,
|
| 16 |
+
"model_specific_special_tokens": {
|
| 17 |
+
"document_token": "<doc>",
|
| 18 |
+
"image_token": "<img>",
|
| 19 |
+
"query_token": "<query>",
|
| 20 |
+
"row_token": "<row>"
|
| 21 |
+
},
|
| 22 |
+
"pad_token": "<pad>",
|
| 23 |
+
"processor_class": "NeoMMEProcessor",
|
| 24 |
+
"query_token": "<query>",
|
| 25 |
+
"row_token": "<row>",
|
| 26 |
+
"tokenizer_class": "TokenizersBackend",
|
| 27 |
+
"unk_token": "<unk>"
|
| 28 |
+
}
|