Upload LlavaNextForConditionalGeneration
Browse files- config.json +5 -4
- generation_config.json +1 -1
- model-00001-of-00002.safetensors +2 -2
- model.safetensors.index.json +1 -1
config.json
CHANGED
|
@@ -3,7 +3,6 @@
|
|
| 3 |
"architectures": [
|
| 4 |
"LlavaNextForConditionalGeneration"
|
| 5 |
],
|
| 6 |
-
"ignore_index": -100,
|
| 7 |
"image_grid_pinpoints": [
|
| 8 |
[
|
| 9 |
384,
|
|
@@ -117,6 +116,7 @@
|
|
| 117 |
"image_seq_length": 576,
|
| 118 |
"image_token_index": 49155,
|
| 119 |
"model_type": "llava_next",
|
|
|
|
| 120 |
"projector_hidden_act": "gelu",
|
| 121 |
"text_config": {
|
| 122 |
"architectures": [
|
|
@@ -139,12 +139,12 @@
|
|
| 139 |
"rms_norm_eps": 1e-05,
|
| 140 |
"rope_theta": 300000,
|
| 141 |
"tie_word_embeddings": true,
|
| 142 |
-
"torch_dtype": "
|
| 143 |
"vocab_size": 49156
|
| 144 |
},
|
| 145 |
"tie_word_embeddings": true,
|
| 146 |
"torch_dtype": "float16",
|
| 147 |
-
"transformers_version": "4.
|
| 148 |
"use_image_newline_parameter": true,
|
| 149 |
"vision_config": {
|
| 150 |
"hidden_act": "gelu_pytorch_tanh",
|
|
@@ -155,7 +155,8 @@
|
|
| 155 |
"model_type": "siglip_vision_model",
|
| 156 |
"num_attention_heads": 16,
|
| 157 |
"num_hidden_layers": 27,
|
| 158 |
-
"patch_size": 14
|
|
|
|
| 159 |
},
|
| 160 |
"vision_feature_layer": [
|
| 161 |
-24,
|
|
|
|
| 3 |
"architectures": [
|
| 4 |
"LlavaNextForConditionalGeneration"
|
| 5 |
],
|
|
|
|
| 6 |
"image_grid_pinpoints": [
|
| 7 |
[
|
| 8 |
384,
|
|
|
|
| 116 |
"image_seq_length": 576,
|
| 117 |
"image_token_index": 49155,
|
| 118 |
"model_type": "llava_next",
|
| 119 |
+
"multimodal_projector_bias": true,
|
| 120 |
"projector_hidden_act": "gelu",
|
| 121 |
"text_config": {
|
| 122 |
"architectures": [
|
|
|
|
| 139 |
"rms_norm_eps": 1e-05,
|
| 140 |
"rope_theta": 300000,
|
| 141 |
"tie_word_embeddings": true,
|
| 142 |
+
"torch_dtype": "float32",
|
| 143 |
"vocab_size": 49156
|
| 144 |
},
|
| 145 |
"tie_word_embeddings": true,
|
| 146 |
"torch_dtype": "float16",
|
| 147 |
+
"transformers_version": "4.49.0",
|
| 148 |
"use_image_newline_parameter": true,
|
| 149 |
"vision_config": {
|
| 150 |
"hidden_act": "gelu_pytorch_tanh",
|
|
|
|
| 155 |
"model_type": "siglip_vision_model",
|
| 156 |
"num_attention_heads": 16,
|
| 157 |
"num_hidden_layers": 27,
|
| 158 |
+
"patch_size": 14,
|
| 159 |
+
"torch_dtype": "float32"
|
| 160 |
},
|
| 161 |
"vision_feature_layer": [
|
| 162 |
-24,
|
generation_config.json
CHANGED
|
@@ -3,5 +3,5 @@
|
|
| 3 |
"bos_token_id": 0,
|
| 4 |
"eos_token_id": 0,
|
| 5 |
"pad_token_id": 0,
|
| 6 |
-
"transformers_version": "4.
|
| 7 |
}
|
|
|
|
| 3 |
"bos_token_id": 0,
|
| 4 |
"eos_token_id": 0,
|
| 5 |
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.49.0"
|
| 7 |
}
|
model-00001-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:36d1e4bbb42ed061d6c0b962025759b8416be3f9b5e59d77fff83eebb674b477
|
| 3 |
+
size 4998717000
|
model.safetensors.index.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"metadata": {
|
| 3 |
-
"total_size":
|
| 4 |
},
|
| 5 |
"weight_map": {
|
| 6 |
"image_newline": "model-00001-of-00002.safetensors",
|
|
|
|
| 1 |
{
|
| 2 |
"metadata": {
|
| 3 |
+
"total_size": 5950793856
|
| 4 |
},
|
| 5 |
"weight_map": {
|
| 6 |
"image_newline": "model-00001-of-00002.safetensors",
|