Update tiny models for Ovis2Model

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

chat_template.jinja

@@ -0,0 +1,7 @@
+<|im_start|>system
+You are a helpful assistant.<|im_end|>
+{% for message in messages %}{{'<|im_start|>' + message['role'] + '
+'}}{% if message['content'] is string %}{{ message['content'] }}{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' %}{{ '<image>
+' }}{% elif content['type'] == 'text' %}{{ content['text'] }}{% endif %}{% endfor %}{% endif %}{{'<|im_end|>
+'}}{% endfor %}{% if add_generation_prompt %}{{'<|im_start|>assistant
+' }}{% endif %}

config.json

@@ -0,0 +1,76 @@
+{
+  "architectures": [
+    "Ovis2Model"
+  ],
+  "dtype": "float32",
+  "hidden_size": 64,
+  "image_token_id": 1,
+  "model_type": "ovis2",
+  "text_config": {
+    "attention_dropout": 0.0,
+    "bos_token_id": null,
+    "eos_token_id": null,
+    "hidden_act": "gelu",
+    "hidden_size": 64,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1",
+      "2": "LABEL_2"
+    },
+    "initializer_range": 0.02,
+    "intermediate_size": 54,
+    "is_training": true,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1,
+      "LABEL_2": 2
+    },
+    "layer_types": [
+      "full_attention",
+      "full_attention"
+    ],
+    "max_position_embeddings": 580,
+    "max_window_layers": 28,
+    "model_type": "qwen2",
+    "num_attention_heads": 4,
+    "num_hidden_layers": 2,
+    "num_key_value_heads": 4,
+    "pad_token_id": 151643,
+    "rms_norm_eps": 1e-06,
+    "rope_parameters": {
+      "rope_theta": 10000.0,
+      "rope_type": "default"
+    },
+    "seq_length": 7,
+    "sliding_window": null,
+    "tie_word_embeddings": false,
+    "use_cache": true,
+    "use_labels": true,
+    "use_sliding_window": false,
+    "vocab_size": 151671
+  },
+  "tie_word_embeddings": true,
+  "transformers_version": "5.16.0.dev0",
+  "vision_config": {
+    "attention_dropout": 0.0,
+    "hidden_act": "silu",
+    "hidden_size": 64,
+    "hidden_stride": 2,
+    "image_size": 32,
+    "initializer_range": 0.02,
+    "intermediate_size": 54,
+    "mlp_bias": false,
+    "model_type": "",
+    "num_attention_heads": 4,
+    "num_channels": 3,
+    "num_hidden_layers": 2,
+    "num_visual_indicator_tokens": 5,
+    "patch_size": 8,
+    "qkv_bias": false,
+    "rms_norm_eps": 1e-05,
+    "tokenize_function": "softmax",
+    "vocab_size": 99
+  },
+  "visual_indicator_token_ids": [],
+  "vocab_size": 151671
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b544560b292b7bf6569d7b2f5d1eae699e76ead96f8873298a895b576a6b39e2
+size 39442384

preprocessor_config.json

@@ -0,0 +1,31 @@
+{
+  "crop_size": {
+    "height": 32,
+    "width": 32
+  },
+  "crop_to_patches": false,
+  "do_convert_rgb": true,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.48145466,
+    0.4578275,
+    0.40821073
+  ],
+  "image_processor_type": "Ovis2ImageProcessor",
+  "image_std": [
+    0.26862954,
+    0.26130258,
+    0.27577711
+  ],
+  "max_patches": 12,
+  "min_patches": 1,
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 32,
+    "width": 32
+  },
+  "use_covering_area_grid": true
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d6666d1c488a64e647514459cb18bab3f80ce5b9b30d3c1d5c53bb06401b4c8
+size 11423012

tokenizer_config.json

@@ -0,0 +1,25 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<IMG_ATOM>",
+    "<IMG_START>",
+    "<IMG_GRID>",
+    "<IMG_COL>",
+    "<IMG_ROW>",
+    "<IMG_END>"
+  ],
+  "is_local": true,
+  "local_files_only": false,
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "processor_class": "Ovis2Processor",
+  "return_token_type_ids": false,
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}