thanks to unum-cloud ❤

.gitattributes

@@ -33,3 +33,5 @@ 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
+preview-girl.png filter=lfs diff=lfs merge=lfs -text
+preview-interior.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,111 @@
+---
+pipeline_tag: image-to-text
+tags:
+  - image-captioning
+  - visual-question-answering
+datasets:
+  - sbu_captions
+  - visual_genome
+  - HuggingFaceM4/VQAv2
+  - ChristophSchuhmann/MS_COCO_2017_URL_TEXT
+language:
+  - en
+license: apache-2.0
+base_model: unum-cloud/uform-vl-english
+widget:
+  - src: preview-interior.png
+    output:
+      text: "The living room is cozy, featuring a red leather chair and a white table. The chair is in the center, and the table is on the left side. A lamp on the left side illuminates the space. A large picture hangs on the wall, adding artistic flair. A vase on the table adds a decorative touch. The room is well-lit, creating a warm and inviting atmosphere."
+  - src: preview-girl.png
+    output:
+      text: "A young girl stands in a grassy field, holding an umbrella to shield herself from the rain. She dons a yellow dress and seems to relish her time outdoors. The umbrella is open, offering protection from the rain. The field is bordered by trees, fostering a tranquil and natural ambiance"
+---
+<Gallery />
+
+<h1 align="center">UForm</h1>
+<h3 align="center">
+Pocket-Sized Multimodal AI<br/>
+For Content Understanding and Generation<br/>
+</h3>
+
+## Description 
+
+UForm-Gen is a small generative vision-language model primarily designed for Image Captioning and Visual Question Answering. The model consists of two parts: 
+
+1. [`uform-vl-english`](https://huggingface.co/unum-cloud/uform-vl-english) visual encoder,
+2. [`Sheared-LLaMA-1.3B`](https://huggingface.co/princeton-nlp/Sheared-LLaMA-1.3B) language model tuned on instruction datasets.
+
+The model was pre-trained on: MSCOCO, SBU Captions, Visual Genome, VQAv2, GQA and a few internal datasets.
+
+### Usage
+
+```bash
+pip install uform
+```
+
+The generative model can be used to caption images, summarize their content, or answer questions about them.
+The exact behavior is controlled by prompts.
+
+```python
+from uform.gen_model import VLMForCausalLM, VLMProcessor
+
+model = VLMForCausalLM.from_pretrained("unum-cloud/uform-gen")
+processor = VLMProcessor.from_pretrained("unum-cloud/uform-gen")
+
+# [cap] Narrate the contents of the image with precision.
+# [cap] Summarize the visual content of the image.
+# [vqa] What is the main subject of the image?
+prompt = "[cap] Summarize the visual content of the image."
+image = Image.open("zebra.jpg")
+
+inputs = processor(texts=[prompt], images=[image], return_tensors="pt")
+with torch.inference_mode():
+     output = model.generate(
+        **inputs,
+        do_sample=False,
+        use_cache=True,
+        max_new_tokens=128,
+        eos_token_id=32001,
+        pad_token_id=processor.tokenizer.pad_token_id
+    )
+
+prompt_len = inputs["input_ids"].shape[1]
+decoded_text = processor.batch_decode(output[:, prompt_len:])[0]
+```
+
+
+## Evaluation
+
+For captioning evaluation we measure CLIPScore and RefCLIPScore¹.
+
+| Model                               | Size | Caption Length | CLIPScore | RefCLIPScore |
+| :---------------------------------- | ---: | -------------: | --------: | -----------: |
+| `llava-hf/llava-1.5-7b-hf`          |   7B |           Long |     0.878 |        0.529 |
+| `llava-hf/llava-1.5-7b-hf`          |   7B |          Short |     0.886 |        0.531 |
+|                                     |
+| `Salesforce/instructblip-vicuna-7b` |   7B |           Long |     0.902 |        0.534 |
+| `Salesforce/instructblip-vicuna-7b` |   7B |          Short |     0.848 |        0.523 |
+|                                     |                                                  |
+| `unum-cloud/uform-gen`              | 1.5B |           Long |     0.847 |        0.523 |
+| `unum-cloud/uform-gen`              | 1.5B |          Short |     0.842 |        0.522 |
+
+Results for VQAv2 evaluation.
+
+| Model                      | Size | Accuracy |
+| :------------------------- | ---: | -------: |
+| `llava-hf/llava-1.5-7b-hf` |   7B |     78.5 |
+| `unum-cloud/uform-gen`     | 1.5B |     66.5 |
+
+¹ We used `apple/DFN5B-CLIP-ViT-H-14-378` CLIP model.
+
+
+## Speed
+
+On RTX 3090, the following performance is expected on text token generation using `float16`, equivalent PyTorch settings, and greedy decoding.
+
+| Model                               | Size |               Speed |   Speedup |
+| :---------------------------------- | ---: | ------------------: | --------: |
+| `llava-hf/llava-1.5-7b-hf`          |   7B |  ~ 40 tokens/second |           |
+| `Salesforce/instructblip-vicuna-7b` |   7B |  ~ 40 tokens/second |           |
+| `unum-cloud/uform-gen`              | 1.5B | ~ 140 tokens/second | __x 3.5__ |
+

added_tokens.json

@@ -0,0 +1,5 @@
+{
+  "<image>": 32002,
+  "<|im_end|>": 32001,
+  "<|im_start|>": 32000
+}

config.json

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "uform-gen",
+  "architectures": [
+    "VLMForCausalLM"
+  ],
+  "center_crop": true,
+  "image_encoder_embedding_dim": 256,
+  "image_encoder_hidden_size": 768,
+  "image_encoder_num_heads": 12,
+  "image_encoder_num_layers": 12,
+  "image_encoder_patch_size": 16,
+  "image_encoder_pooling": "cls",
+  "image_pooler_intermediate_size": 5504,
+  "image_pooler_num_attn_heads": 16,
+  "image_pooler_num_latents": 196,
+  "image_size": 224,
+  "image_token_id": 32002,
+  "initializer_range": 0.02,
+  "model_type": "vlm",
+  "text_decoder_name_or_path": "princeton-nlp/Sheared-LLaMA-1.3B",
+  "tokenizer_name_or_path": "unum-cloud/uform-gen",
+  "torch_dtype": "float32",
+  "transformers_version": "4.36.2",
+  "use_cache": true
+}

generation_config.json

@@ -0,0 +1,4 @@
+{
+  "_from_model_config": true,
+  "transformers_version": "4.36.2"
+}

model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c7ae42a70765c8dc2d47ed8c5c2cdd7ced413b6d08b58f8983d639b96be87f4b
+size 4984310240

model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1bff12925d06fc2207053223358a13175d0b698a78b947154182676266dda7f
+size 973984040

model.safetensors.index.json

@@ -0,0 +1,470 @@
+{
+  "metadata": {
+    "total_size": 5958239744
+  },
+  "weight_map": {
+    "image_encoder.blocks.0.attn.key.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.attn.key.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.attn.out.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.attn.out.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.attn.query.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.attn.query.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.attn.value.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.attn.value.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.ls1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.ls2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.mlp.hidden_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.mlp.hidden_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.mlp.output_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.mlp.output_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.0.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.attn.key.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.attn.key.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.attn.out.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.attn.out.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.attn.query.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.attn.query.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.attn.value.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.attn.value.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.ls1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.ls2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.mlp.hidden_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.mlp.hidden_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.mlp.output_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.mlp.output_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.1.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.attn.key.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.attn.key.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.attn.out.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.attn.out.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.attn.query.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.attn.query.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.attn.value.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.attn.value.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.ls1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.ls2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.mlp.hidden_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.mlp.hidden_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.mlp.output_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.mlp.output_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.10.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.attn.key.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.attn.key.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.attn.out.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.attn.out.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.attn.query.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.attn.query.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.attn.value.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.attn.value.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.ls1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.ls2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.mlp.hidden_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.mlp.hidden_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.mlp.output_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.mlp.output_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.11.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.attn.key.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.attn.key.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.attn.out.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.attn.out.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.attn.query.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.attn.query.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.attn.value.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.attn.value.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.ls1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.ls2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.mlp.hidden_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.mlp.hidden_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.mlp.output_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.mlp.output_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.2.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.attn.key.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.attn.key.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.attn.out.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.attn.out.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.attn.query.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.attn.query.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.attn.value.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.attn.value.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.ls1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.ls2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.mlp.hidden_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.mlp.hidden_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.mlp.output_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.mlp.output_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.3.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.attn.key.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.attn.key.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.attn.out.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.attn.out.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.attn.query.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.attn.query.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.attn.value.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.attn.value.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.ls1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.ls2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.mlp.hidden_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.mlp.hidden_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.mlp.output_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.mlp.output_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.4.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.attn.key.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.attn.key.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.attn.out.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.attn.out.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.attn.query.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.attn.query.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.attn.value.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.attn.value.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.ls1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.ls2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.mlp.hidden_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.mlp.hidden_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.mlp.output_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.mlp.output_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.5.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.attn.key.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.attn.key.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.attn.out.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.attn.out.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.attn.query.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.attn.query.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.attn.value.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.attn.value.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.ls1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.ls2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.mlp.hidden_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.mlp.hidden_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.mlp.output_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.mlp.output_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.6.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.attn.key.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.attn.key.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.attn.out.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.attn.out.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.attn.query.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.attn.query.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.attn.value.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.attn.value.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.ls1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.ls2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.mlp.hidden_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.mlp.hidden_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.mlp.output_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.mlp.output_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.7.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.attn.key.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.attn.key.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.attn.out.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.attn.out.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.attn.query.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.attn.query.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.attn.value.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.attn.value.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.ls1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.ls2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.mlp.hidden_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.mlp.hidden_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.mlp.output_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.mlp.output_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.8.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.attn.key.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.attn.key.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.attn.out.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.attn.out.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.attn.query.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.attn.query.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.attn.value.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.attn.value.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.ls1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.ls2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.mlp.hidden_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.mlp.hidden_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.mlp.output_layer.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.mlp.output_layer.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.blocks.9.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.cls_token": "model-00002-of-00002.safetensors",
+    "image_encoder.embedding_projection.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.norm.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.norm.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.patch_embed.bias": "model-00002-of-00002.safetensors",
+    "image_encoder.patch_embed.weight": "model-00002-of-00002.safetensors",
+    "image_encoder.pos_embed": "model-00002-of-00002.safetensors",
+    "image_pooler.image_latents": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.linear1.bias": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.linear1.weight": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.linear2.bias": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.linear2.weight": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.multihead_attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.multihead_attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.multihead_attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.multihead_attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.norm1.bias": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.norm1.weight": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.norm2.bias": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.norm2.weight": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.norm3.bias": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.norm3.weight": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.self_attn.in_proj_bias": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.self_attn.in_proj_weight": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.self_attn.out_proj.bias": "model-00002-of-00002.safetensors",
+    "image_pooler.pooler.self_attn.out_proj.weight": "model-00002-of-00002.safetensors",
+    "image_pooler.projection.bias": "model-00002-of-00002.safetensors",
+    "image_pooler.projection.weight": "model-00002-of-00002.safetensors",
+    "text_decoder.lm_head.weight": "model-00002-of-00002.safetensors",
+    "text_decoder.model.embed_tokens.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors",
+    "text_decoder.model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+    "text_decoder.model.layers.23.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+    "text_decoder.model.layers.23.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+    "text_decoder.model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
+    "text_decoder.model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+    "text_decoder.model.norm.weight": "model-00002-of-00002.safetensors"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,30 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

tokenizer_config.json

@@ -0,0 +1,68 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "32000": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "32001": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "32002": {
+      "content": "<image>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "bos_token": "<s>",
+  "chat_template": "{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% for message in messages %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + ' <|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "max_length": 2048,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "</s>",
+  "sp_model_kwargs": {},
+  "stride": 0,
+  "tokenizer_class": "LlamaTokenizer",
+  "truncation_side": "left",
+  "truncation_strategy": "longest_first",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}