Upload CXRMate2ForConditionalGeneration
Browse files- config.json +5 -117
- generation_config.json +1 -1
- model.safetensors.index.json +1 -0
config.json
CHANGED
|
@@ -7,6 +7,7 @@
|
|
| 7 |
"AutoModelForCausalLM": "modelling_cxrmate2.CXRMate2ForConditionalGeneration"
|
| 8 |
},
|
| 9 |
"bos_token_id": 128000,
|
|
|
|
| 10 |
"findings_token_type_id": 128029,
|
| 11 |
"generate_both_sections_token_id": null,
|
| 12 |
"generate_findings_token_id": 128003,
|
|
@@ -25,70 +26,28 @@
|
|
| 25 |
"projector_hidden_act": null,
|
| 26 |
"sep_token_id": 128072,
|
| 27 |
"text_config": {
|
| 28 |
-
"_attn_implementation_autoset": false,
|
| 29 |
"_name_or_path": "meta-llama/Llama-3.2-3B",
|
| 30 |
-
"add_cross_attention": false,
|
| 31 |
"architectures": [
|
| 32 |
"LlamaForCausalLM"
|
| 33 |
],
|
| 34 |
"attention_bias": false,
|
| 35 |
"attention_dropout": 0.0,
|
| 36 |
"attn_implementation": "eager",
|
| 37 |
-
"bad_words_ids": null,
|
| 38 |
-
"begin_suppress_tokens": null,
|
| 39 |
"bos_token_id": 128000,
|
| 40 |
-
"
|
| 41 |
-
"cross_attention_hidden_size": null,
|
| 42 |
-
"decoder_start_token_id": null,
|
| 43 |
-
"diversity_penalty": 0.0,
|
| 44 |
-
"do_sample": false,
|
| 45 |
-
"early_stopping": false,
|
| 46 |
-
"encoder_no_repeat_ngram_size": 0,
|
| 47 |
"eos_token_id": 128001,
|
| 48 |
-
"exponential_decay_length_penalty": null,
|
| 49 |
-
"finetuning_task": null,
|
| 50 |
-
"forced_bos_token_id": null,
|
| 51 |
-
"forced_eos_token_id": null,
|
| 52 |
"head_dim": 128,
|
| 53 |
"hidden_act": "silu",
|
| 54 |
"hidden_size": 3072,
|
| 55 |
-
"id2label": {
|
| 56 |
-
"0": "LABEL_0",
|
| 57 |
-
"1": "LABEL_1"
|
| 58 |
-
},
|
| 59 |
"initializer_range": 0.02,
|
| 60 |
"intermediate_size": 8192,
|
| 61 |
-
"is_decoder": false,
|
| 62 |
-
"is_encoder_decoder": false,
|
| 63 |
-
"label2id": {
|
| 64 |
-
"LABEL_0": 0,
|
| 65 |
-
"LABEL_1": 1
|
| 66 |
-
},
|
| 67 |
-
"length_penalty": 1.0,
|
| 68 |
-
"max_length": 20,
|
| 69 |
"max_position_embeddings": 131072,
|
| 70 |
-
"min_length": 0,
|
| 71 |
"mlp_bias": false,
|
| 72 |
"model_type": "llama",
|
| 73 |
-
"no_repeat_ngram_size": 0,
|
| 74 |
"num_attention_heads": 24,
|
| 75 |
-
"num_beam_groups": 1,
|
| 76 |
-
"num_beams": 1,
|
| 77 |
"num_hidden_layers": 28,
|
| 78 |
"num_key_value_heads": 8,
|
| 79 |
-
"num_return_sequences": 1,
|
| 80 |
-
"output_attentions": false,
|
| 81 |
-
"output_hidden_states": false,
|
| 82 |
-
"output_scores": false,
|
| 83 |
-
"pad_token_id": null,
|
| 84 |
-
"prefix": null,
|
| 85 |
"pretraining_tp": 1,
|
| 86 |
-
"problem_type": null,
|
| 87 |
-
"pruned_heads": {},
|
| 88 |
-
"remove_invalid_values": false,
|
| 89 |
-
"repetition_penalty": 1.0,
|
| 90 |
-
"return_dict": true,
|
| 91 |
-
"return_dict_in_generate": false,
|
| 92 |
"rms_norm_eps": 1e-05,
|
| 93 |
"rope_scaling": {
|
| 94 |
"factor": 32.0,
|
|
@@ -98,103 +57,44 @@
|
|
| 98 |
"rope_type": "llama3"
|
| 99 |
},
|
| 100 |
"rope_theta": 500000.0,
|
| 101 |
-
"sep_token_id": null,
|
| 102 |
-
"suppress_tokens": null,
|
| 103 |
-
"task_specific_params": null,
|
| 104 |
-
"temperature": 1.0,
|
| 105 |
-
"tf_legacy_loss": false,
|
| 106 |
-
"tie_encoder_decoder": false,
|
| 107 |
"tie_word_embeddings": true,
|
| 108 |
-
"tokenizer_class": null,
|
| 109 |
-
"top_k": 50,
|
| 110 |
-
"top_p": 1.0,
|
| 111 |
-
"torch_dtype": "bfloat16",
|
| 112 |
-
"torchscript": false,
|
| 113 |
-
"typical_p": 1.0,
|
| 114 |
-
"use_bfloat16": false,
|
| 115 |
"use_cache": true,
|
| 116 |
"vocab_size": 128256
|
| 117 |
},
|
| 118 |
"time_delta_encoder_intermediate_size": 2048,
|
| 119 |
"time_delta_monotonic_inversion": true,
|
| 120 |
-
"
|
| 121 |
-
"transformers_version": "4.47.1",
|
| 122 |
"vision_config": {
|
| 123 |
-
"_attn_implementation_autoset": false,
|
| 124 |
"_name_or_path": "microsoft/rad-dino-maira-2",
|
| 125 |
-
"add_cross_attention": false,
|
| 126 |
"apply_layernorm": true,
|
| 127 |
"architectures": [
|
| 128 |
"Dinov2Model"
|
| 129 |
],
|
| 130 |
"attention_probs_dropout_prob": 0.0,
|
| 131 |
"attn_implementation": "eager",
|
| 132 |
-
"bad_words_ids": null,
|
| 133 |
-
"begin_suppress_tokens": null,
|
| 134 |
-
"bos_token_id": null,
|
| 135 |
-
"chunk_size_feed_forward": 0,
|
| 136 |
-
"cross_attention_hidden_size": null,
|
| 137 |
-
"decoder_start_token_id": null,
|
| 138 |
-
"diversity_penalty": 0.0,
|
| 139 |
-
"do_sample": false,
|
| 140 |
"drop_path_rate": 0.0,
|
| 141 |
-
"
|
| 142 |
-
"encoder_no_repeat_ngram_size": 0,
|
| 143 |
-
"eos_token_id": null,
|
| 144 |
-
"exponential_decay_length_penalty": null,
|
| 145 |
-
"finetuning_task": null,
|
| 146 |
-
"forced_bos_token_id": null,
|
| 147 |
-
"forced_eos_token_id": null,
|
| 148 |
"hidden_act": "gelu",
|
| 149 |
"hidden_dropout_prob": 0.0,
|
| 150 |
"hidden_size": 768,
|
| 151 |
-
"id2label": {
|
| 152 |
-
"0": "LABEL_0",
|
| 153 |
-
"1": "LABEL_1"
|
| 154 |
-
},
|
| 155 |
"image_size": 518,
|
| 156 |
"initializer_range": 0.02,
|
| 157 |
-
"is_decoder": false,
|
| 158 |
-
"is_encoder_decoder": false,
|
| 159 |
-
"label2id": {
|
| 160 |
-
"LABEL_0": 0,
|
| 161 |
-
"LABEL_1": 1
|
| 162 |
-
},
|
| 163 |
"layer_norm_eps": 1e-06,
|
| 164 |
"layerscale_value": 1.0,
|
| 165 |
-
"length_penalty": 1.0,
|
| 166 |
-
"max_length": 20,
|
| 167 |
-
"min_length": 0,
|
| 168 |
"mlp_ratio": 4,
|
| 169 |
"model_type": "dinov2",
|
| 170 |
-
"no_repeat_ngram_size": 0,
|
| 171 |
"num_attention_heads": 12,
|
| 172 |
-
"num_beam_groups": 1,
|
| 173 |
-
"num_beams": 1,
|
| 174 |
"num_channels": 3,
|
| 175 |
"num_hidden_layers": 12,
|
| 176 |
-
"num_return_sequences": 1,
|
| 177 |
"out_features": [
|
| 178 |
"stage12"
|
| 179 |
],
|
| 180 |
"out_indices": [
|
| 181 |
12
|
| 182 |
],
|
| 183 |
-
"output_attentions": false,
|
| 184 |
-
"output_hidden_states": false,
|
| 185 |
-
"output_scores": false,
|
| 186 |
-
"pad_token_id": null,
|
| 187 |
"patch_size": 14,
|
| 188 |
-
"prefix": null,
|
| 189 |
-
"problem_type": null,
|
| 190 |
-
"pruned_heads": {},
|
| 191 |
"qkv_bias": true,
|
| 192 |
-
"remove_invalid_values": false,
|
| 193 |
-
"repetition_penalty": 1.0,
|
| 194 |
"reshape_hidden_states": true,
|
| 195 |
-
"return_dict": true,
|
| 196 |
-
"return_dict_in_generate": false,
|
| 197 |
-
"sep_token_id": null,
|
| 198 |
"stage_names": [
|
| 199 |
"stem",
|
| 200 |
"stage1",
|
|
@@ -210,19 +110,7 @@
|
|
| 210 |
"stage11",
|
| 211 |
"stage12"
|
| 212 |
],
|
| 213 |
-
"
|
| 214 |
-
"task_specific_params": null,
|
| 215 |
-
"temperature": 1.0,
|
| 216 |
-
"tf_legacy_loss": false,
|
| 217 |
-
"tie_encoder_decoder": false,
|
| 218 |
-
"tie_word_embeddings": true,
|
| 219 |
-
"tokenizer_class": null,
|
| 220 |
-
"top_k": 50,
|
| 221 |
-
"top_p": 1.0,
|
| 222 |
-
"torch_dtype": "float32",
|
| 223 |
-
"torchscript": false,
|
| 224 |
-
"typical_p": 1.0,
|
| 225 |
-
"use_bfloat16": false,
|
| 226 |
"use_swiglu_ffn": false
|
| 227 |
},
|
| 228 |
"vision_feature_layer": -1,
|
|
|
|
| 7 |
"AutoModelForCausalLM": "modelling_cxrmate2.CXRMate2ForConditionalGeneration"
|
| 8 |
},
|
| 9 |
"bos_token_id": 128000,
|
| 10 |
+
"dtype": "float32",
|
| 11 |
"findings_token_type_id": 128029,
|
| 12 |
"generate_both_sections_token_id": null,
|
| 13 |
"generate_findings_token_id": 128003,
|
|
|
|
| 26 |
"projector_hidden_act": null,
|
| 27 |
"sep_token_id": 128072,
|
| 28 |
"text_config": {
|
|
|
|
| 29 |
"_name_or_path": "meta-llama/Llama-3.2-3B",
|
|
|
|
| 30 |
"architectures": [
|
| 31 |
"LlamaForCausalLM"
|
| 32 |
],
|
| 33 |
"attention_bias": false,
|
| 34 |
"attention_dropout": 0.0,
|
| 35 |
"attn_implementation": "eager",
|
|
|
|
|
|
|
| 36 |
"bos_token_id": 128000,
|
| 37 |
+
"dtype": "float32",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
"eos_token_id": 128001,
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
"head_dim": 128,
|
| 40 |
"hidden_act": "silu",
|
| 41 |
"hidden_size": 3072,
|
|
|
|
|
|
|
|
|
|
|
|
|
| 42 |
"initializer_range": 0.02,
|
| 43 |
"intermediate_size": 8192,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
"max_position_embeddings": 131072,
|
|
|
|
| 45 |
"mlp_bias": false,
|
| 46 |
"model_type": "llama",
|
|
|
|
| 47 |
"num_attention_heads": 24,
|
|
|
|
|
|
|
| 48 |
"num_hidden_layers": 28,
|
| 49 |
"num_key_value_heads": 8,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
"pretraining_tp": 1,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 51 |
"rms_norm_eps": 1e-05,
|
| 52 |
"rope_scaling": {
|
| 53 |
"factor": 32.0,
|
|
|
|
| 57 |
"rope_type": "llama3"
|
| 58 |
},
|
| 59 |
"rope_theta": 500000.0,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 60 |
"tie_word_embeddings": true,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 61 |
"use_cache": true,
|
| 62 |
"vocab_size": 128256
|
| 63 |
},
|
| 64 |
"time_delta_encoder_intermediate_size": 2048,
|
| 65 |
"time_delta_monotonic_inversion": true,
|
| 66 |
+
"transformers_version": "4.57.1",
|
|
|
|
| 67 |
"vision_config": {
|
|
|
|
| 68 |
"_name_or_path": "microsoft/rad-dino-maira-2",
|
|
|
|
| 69 |
"apply_layernorm": true,
|
| 70 |
"architectures": [
|
| 71 |
"Dinov2Model"
|
| 72 |
],
|
| 73 |
"attention_probs_dropout_prob": 0.0,
|
| 74 |
"attn_implementation": "eager",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 75 |
"drop_path_rate": 0.0,
|
| 76 |
+
"dtype": "float32",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 77 |
"hidden_act": "gelu",
|
| 78 |
"hidden_dropout_prob": 0.0,
|
| 79 |
"hidden_size": 768,
|
|
|
|
|
|
|
|
|
|
|
|
|
| 80 |
"image_size": 518,
|
| 81 |
"initializer_range": 0.02,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 82 |
"layer_norm_eps": 1e-06,
|
| 83 |
"layerscale_value": 1.0,
|
|
|
|
|
|
|
|
|
|
| 84 |
"mlp_ratio": 4,
|
| 85 |
"model_type": "dinov2",
|
|
|
|
| 86 |
"num_attention_heads": 12,
|
|
|
|
|
|
|
| 87 |
"num_channels": 3,
|
| 88 |
"num_hidden_layers": 12,
|
|
|
|
| 89 |
"out_features": [
|
| 90 |
"stage12"
|
| 91 |
],
|
| 92 |
"out_indices": [
|
| 93 |
12
|
| 94 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
| 95 |
"patch_size": 14,
|
|
|
|
|
|
|
|
|
|
| 96 |
"qkv_bias": true,
|
|
|
|
|
|
|
| 97 |
"reshape_hidden_states": true,
|
|
|
|
|
|
|
|
|
|
| 98 |
"stage_names": [
|
| 99 |
"stem",
|
| 100 |
"stage1",
|
|
|
|
| 110 |
"stage11",
|
| 111 |
"stage12"
|
| 112 |
],
|
| 113 |
+
"use_mask_token": true,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 114 |
"use_swiglu_ffn": false
|
| 115 |
},
|
| 116 |
"vision_feature_layer": -1,
|
generation_config.json
CHANGED
|
@@ -2,5 +2,5 @@
|
|
| 2 |
"_from_model_config": true,
|
| 3 |
"bos_token_id": 128000,
|
| 4 |
"eos_token_id": 128001,
|
| 5 |
-
"transformers_version": "4.
|
| 6 |
}
|
|
|
|
| 2 |
"_from_model_config": true,
|
| 3 |
"bos_token_id": 128000,
|
| 4 |
"eos_token_id": 128001,
|
| 5 |
+
"transformers_version": "4.57.1"
|
| 6 |
}
|
model.safetensors.index.json
CHANGED
|
@@ -1,5 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"metadata": {
|
|
|
|
| 3 |
"total_size": 13289040896
|
| 4 |
},
|
| 5 |
"weight_map": {
|
|
|
|
| 1 |
{
|
| 2 |
"metadata": {
|
| 3 |
+
"total_parameters": 3322260224,
|
| 4 |
"total_size": 13289040896
|
| 5 |
},
|
| 6 |
"weight_map": {
|