Initial MLX conversion of docling-project/CodeFormulaV2

README.md

@@ -0,0 +1,59 @@
+---
+language: en
+license: cdla-permissive-2.0
+pipeline_tag: image-text-to-text
+library_name: mlx
+tags:
+  - mlx
+  - mlx-vlm
+  - idefics3
+base_model: docling-project/CodeFormulaV2
+base_model_relation: quantized
+datasets:
+  - ds4sd/SynthFormulaNet
+  - ds4sd/SynthCodeNet
+---
+
+# CodeFormulaV2-mlx-bf16
+
+MLX bf16 conversion of [docling-project/CodeFormulaV2](https://huggingface.co/docling-project/CodeFormulaV2), produced with [`mlx_vlm.convert`](https://github.com/Blaizzy/mlx-vlm). Architecture, training data, and intended use are described on the upstream model page; this repo only changes the storage format (PyTorch safetensors → MLX safetensors, same bf16 precision).
+
+## Usage
+
+```bash
+pip install mlx-vlm
+```
+
+```python
+from mlx_vlm import load, generate
+from mlx_vlm.prompt_utils import apply_chat_template
+
+model, processor = load("mlx-community/CodeFormulaV2-mlx-bf16")
+prompt = apply_chat_template(processor, model.config, "<formula>", num_images=1)
+result = generate(
+    model, processor,
+    prompt=prompt,
+    image="path/to/image.png",
+    temperature=0.0,
+)
+print(result.text)
+```
+
+Use `"<formula>"` as the prompt for a math-expression image, `"<code>"` for a code-block image, per the upstream model card.
+
+## License and attribution
+
+This is a derivative of [docling-project/CodeFormulaV2](https://huggingface.co/docling-project/CodeFormulaV2), redistributed under the same [Community Data License Agreement – Permissive 2.0 (CDLA-Permissive-2.0)](https://cdla.dev/permissive-2-0/).
+
+Please cite the upstream work:
+
+```bibtex
+@techreport{Docling,
+  author = {Deep Search Team},
+  month = {8},
+  title = {{Docling Technical Report}},
+  url = {https://arxiv.org/abs/2408.09869},
+  eprint = {2408.09869},
+  year = {2024}
+}
+```

added_tokens.json

@@ -0,0 +1,38 @@
+{
+  "<row_1_col_1>": 100352,
+  "<row_1_col_2>": 100353,
+  "<row_1_col_3>": 100354,
+  "<row_1_col_4>": 100355,
+  "<row_1_col_5>": 100356,
+  "<row_1_col_6>": 100357,
+  "<row_2_col_1>": 100358,
+  "<row_2_col_2>": 100359,
+  "<row_2_col_3>": 100360,
+  "<row_2_col_4>": 100361,
+  "<row_2_col_5>": 100362,
+  "<row_2_col_6>": 100363,
+  "<row_3_col_1>": 100364,
+  "<row_3_col_2>": 100365,
+  "<row_3_col_3>": 100366,
+  "<row_3_col_4>": 100367,
+  "<row_3_col_5>": 100368,
+  "<row_3_col_6>": 100369,
+  "<row_4_col_1>": 100370,
+  "<row_4_col_2>": 100371,
+  "<row_4_col_3>": 100372,
+  "<row_4_col_4>": 100373,
+  "<row_4_col_5>": 100374,
+  "<row_4_col_6>": 100375,
+  "<row_5_col_1>": 100376,
+  "<row_5_col_2>": 100377,
+  "<row_5_col_3>": 100378,
+  "<row_5_col_4>": 100379,
+  "<row_5_col_5>": 100380,
+  "<row_5_col_6>": 100381,
+  "<row_6_col_1>": 100382,
+  "<row_6_col_2>": 100383,
+  "<row_6_col_3>": 100384,
+  "<row_6_col_4>": 100385,
+  "<row_6_col_5>": 100386,
+  "<row_6_col_6>": 100387
+}

chat_template.jinja

@@ -0,0 +1,2 @@
+<|start_of_role|>{% for message in messages %}{{message['role']}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>
+{% endfor %}{% if add_generation_prompt %}{{ 'assistant:' }}{% endif %}

chat_template.json

@@ -0,0 +1,3 @@
+{
+  "chat_template": "<|start_of_role|>{% for message in messages %}{{message['role']}}{% if message['content'][0]['type'] == 'image' %}{{':'}}{% else %}{{': '}}{% endif %}{% for line in message['content'] %}{% if line['type'] == 'text' %}{{line['text']}}{% elif line['type'] == 'image' %}{{ '<image>' }}{% endif %}{% endfor %}<end_of_utterance>\n{% endfor %}{% if add_generation_prompt %}{{ 'assistant:' }}{% endif %}"
+}

config.json

@@ -0,0 +1,102 @@
+{
+    "_flash_attn_2_enabled": true,
+    "architectures": [
+        "Idefics3ForConditionalGeneration"
+    ],
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "bos_token_id": 100264,
+    "eos_token_id": 100338,
+    "freeze_lm_head": true,
+    "freeze_text_layers": true,
+    "freeze_text_module_exceptions": [],
+    "freeze_vision_layers": true,
+    "freeze_vision_module_exceptions": [],
+    "head_dim": 64,
+    "hidden_act": "silu",
+    "hidden_size": 576,
+    "image_token_id": 100270,
+    "initializer_range": 0.02,
+    "intermediate_size": 1536,
+    "max_position_embeddings": 8192,
+    "mlp_bias": false,
+    "model_type": "idefics3",
+    "neftune_noise_alpha": 0.0,
+    "num_attention_heads": 9,
+    "num_hidden_layers": 30,
+    "num_key_value_heads": 3,
+    "pad_token_id": 100256,
+    "perceiver_config": {
+        "attention_dropout": 0.0,
+        "hidden_act": "silu",
+        "model_type": "vllama3",
+        "num_key_value_heads": 1,
+        "qk_layer_norms_perceiver": false,
+        "resampler_depth": 6,
+        "resampler_head_dim": 96,
+        "resampler_n_heads": 16,
+        "resampler_n_latents": 64
+    },
+    "pixel_shuffle_factor": 4,
+    "pretraining_tp": 1,
+    "qk_layer_norms": false,
+    "rms_norm_eps": 1e-05,
+    "rope_scaling": null,
+    "rope_theta": 100000.0,
+    "scale_factor": 4,
+    "text_config": {
+        "architectures": [
+            "VLlama3ForCausalLM"
+        ],
+        "attention_bias": false,
+        "attention_dropout": 0.0,
+        "bos_token_id": 100257,
+        "eos_token_id": 100257,
+        "head_dim": 64,
+        "hidden_act": "silu",
+        "hidden_size": 576,
+        "initializer_range": 0.02,
+        "intermediate_size": 1536,
+        "max_position_embeddings": 8192,
+        "mlp_bias": false,
+        "model_type": "llama",
+        "num_attention_heads": 9,
+        "num_hidden_layers": 30,
+        "num_key_value_heads": 3,
+        "pretraining_tp": 1,
+        "rms_norm_eps": 1e-05,
+        "rope_scaling": null,
+        "rope_theta": 10000.0,
+        "torch_dtype": "bfloat16",
+        "use_cache": true,
+        "vocab_size": 100480,
+        "tie_word_embeddings": false
+    },
+    "tie_word_embeddings": true,
+    "transformers_version": "4.51.3",
+    "use_cache": false,
+    "use_resampler": false,
+    "vision_config": {
+        "attention_dropout": 0.0,
+        "hidden_act": "gelu_pytorch_tanh",
+        "hidden_size": 768,
+        "image_size": 512,
+        "initializer_range": 0.02,
+        "intermediate_size": 3072,
+        "layer_norm_eps": 1e-06,
+        "max_image_size": {
+            "longest_edge": 512
+        },
+        "model_type": "idefics3_vision",
+        "num_attention_heads": 12,
+        "num_channels": 3,
+        "num_hidden_layers": 12,
+        "patch_size": 16,
+        "size": {
+            "longest_edge": 2048
+        },
+        "torch_dtype": "bfloat16",
+        "use_base_siglip": true
+    },
+    "vocab_size": 100480
+}

generation_config.json

@@ -0,0 +1,8 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 100264,
+  "eos_token_id": 100338,
+  "pad_token_id": 100256,
+  "transformers_version": "4.51.3",
+  "use_cache": false
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3dd26ac59ae5f62a586d40070e7beeda16ba66847146c5314c6b76daf733ea3f
+size 630991636

model.safetensors.index.json

@@ -0,0 +1,478 @@
+{
+    "metadata": {
+        "total_size": 630934656
+    },
+    "weight_map": {
+        "connector.modality_projection.proj.weight": "model.safetensors",
+        "language_model.embed_tokens.weight": "model.safetensors",
+        "language_model.layers.0.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.0.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.0.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.0.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.0.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.0.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.0.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.0.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.0.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.1.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.1.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.1.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.1.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.1.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.1.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.1.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.1.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.1.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.10.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.10.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.10.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.10.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.10.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.10.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.10.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.10.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.10.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.11.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.11.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.11.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.11.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.11.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.11.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.11.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.11.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.11.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.12.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.12.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.12.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.12.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.12.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.12.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.12.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.12.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.12.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.13.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.13.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.13.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.13.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.13.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.13.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.13.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.13.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.13.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.14.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.14.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.14.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.14.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.14.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.14.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.14.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.14.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.14.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.15.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.15.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.15.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.15.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.15.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.15.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.15.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.15.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.15.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.16.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.16.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.16.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.16.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.16.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.16.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.16.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.16.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.16.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.17.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.17.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.17.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.17.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.17.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.17.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.17.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.17.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.17.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.18.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.18.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.18.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.18.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.18.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.18.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.18.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.18.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.18.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.19.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.19.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.19.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.19.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.19.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.19.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.19.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.19.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.19.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.2.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.2.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.2.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.2.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.2.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.2.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.2.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.2.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.2.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.20.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.20.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.20.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.20.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.20.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.20.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.20.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.20.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.20.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.21.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.21.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.21.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.21.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.21.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.21.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.21.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.21.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.21.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.22.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.22.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.22.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.22.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.22.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.22.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.22.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.22.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.22.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.23.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.23.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.23.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.23.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.23.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.23.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.23.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.23.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.23.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.24.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.24.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.24.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.24.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.24.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.24.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.24.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.24.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.24.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.25.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.25.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.25.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.25.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.25.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.25.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.25.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.25.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.25.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.26.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.26.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.26.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.26.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.26.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.26.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.26.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.26.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.26.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.27.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.27.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.27.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.27.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.27.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.27.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.27.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.27.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.27.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.28.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.28.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.28.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.28.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.28.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.28.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.28.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.28.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.28.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.29.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.29.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.29.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.29.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.29.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.29.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.29.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.29.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.29.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.3.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.3.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.3.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.3.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.3.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.3.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.3.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.3.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.3.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.4.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.4.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.4.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.4.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.4.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.4.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.4.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.4.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.4.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.5.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.5.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.5.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.5.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.5.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.5.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.5.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.5.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.5.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.6.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.6.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.6.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.6.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.6.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.6.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.6.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.6.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.6.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.7.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.7.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.7.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.7.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.7.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.7.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.7.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.7.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.7.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.8.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.8.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.8.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.8.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.8.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.8.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.8.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.8.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.8.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.layers.9.input_layernorm.weight": "model.safetensors",
+        "language_model.layers.9.mlp.down_proj.weight": "model.safetensors",
+        "language_model.layers.9.mlp.gate_proj.weight": "model.safetensors",
+        "language_model.layers.9.mlp.up_proj.weight": "model.safetensors",
+        "language_model.layers.9.post_attention_layernorm.weight": "model.safetensors",
+        "language_model.layers.9.self_attn.k_proj.weight": "model.safetensors",
+        "language_model.layers.9.self_attn.o_proj.weight": "model.safetensors",
+        "language_model.layers.9.self_attn.q_proj.weight": "model.safetensors",
+        "language_model.layers.9.self_attn.v_proj.weight": "model.safetensors",
+        "language_model.lm_head.weight": "model.safetensors",
+        "language_model.norm.weight": "model.safetensors",
+        "vision_model.embeddings.patch_embedding.bias": "model.safetensors",
+        "vision_model.embeddings.patch_embedding.weight": "model.safetensors",
+        "vision_model.embeddings.position_embedding.weight": "model.safetensors",
+        "vision_model.encoder.layers.0.layer_norm1.bias": "model.safetensors",
+        "vision_model.encoder.layers.0.layer_norm1.weight": "model.safetensors",
+        "vision_model.encoder.layers.0.layer_norm2.bias": "model.safetensors",
+        "vision_model.encoder.layers.0.layer_norm2.weight": "model.safetensors",
+        "vision_model.encoder.layers.0.mlp.fc1.bias": "model.safetensors",
+        "vision_model.encoder.layers.0.mlp.fc1.weight": "model.safetensors",
+        "vision_model.encoder.layers.0.mlp.fc2.bias": "model.safetensors",
+        "vision_model.encoder.layers.0.mlp.fc2.weight": "model.safetensors",
+        "vision_model.encoder.layers.0.self_attn.k_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.0.self_attn.k_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.0.self_attn.out_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.0.self_attn.out_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.0.self_attn.q_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.0.self_attn.q_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.0.self_attn.v_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.0.self_attn.v_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.1.layer_norm1.bias": "model.safetensors",
+        "vision_model.encoder.layers.1.layer_norm1.weight": "model.safetensors",
+        "vision_model.encoder.layers.1.layer_norm2.bias": "model.safetensors",
+        "vision_model.encoder.layers.1.layer_norm2.weight": "model.safetensors",
+        "vision_model.encoder.layers.1.mlp.fc1.bias": "model.safetensors",
+        "vision_model.encoder.layers.1.mlp.fc1.weight": "model.safetensors",
+        "vision_model.encoder.layers.1.mlp.fc2.bias": "model.safetensors",
+        "vision_model.encoder.layers.1.mlp.fc2.weight": "model.safetensors",
+        "vision_model.encoder.layers.1.self_attn.k_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.1.self_attn.k_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.1.self_attn.out_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.1.self_attn.out_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.1.self_attn.q_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.1.self_attn.q_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.1.self_attn.v_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.1.self_attn.v_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.10.layer_norm1.bias": "model.safetensors",
+        "vision_model.encoder.layers.10.layer_norm1.weight": "model.safetensors",
+        "vision_model.encoder.layers.10.layer_norm2.bias": "model.safetensors",
+        "vision_model.encoder.layers.10.layer_norm2.weight": "model.safetensors",
+        "vision_model.encoder.layers.10.mlp.fc1.bias": "model.safetensors",
+        "vision_model.encoder.layers.10.mlp.fc1.weight": "model.safetensors",
+        "vision_model.encoder.layers.10.mlp.fc2.bias": "model.safetensors",
+        "vision_model.encoder.layers.10.mlp.fc2.weight": "model.safetensors",
+        "vision_model.encoder.layers.10.self_attn.k_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.10.self_attn.k_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.10.self_attn.out_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.10.self_attn.out_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.10.self_attn.q_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.10.self_attn.q_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.10.self_attn.v_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.10.self_attn.v_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.11.layer_norm1.bias": "model.safetensors",
+        "vision_model.encoder.layers.11.layer_norm1.weight": "model.safetensors",
+        "vision_model.encoder.layers.11.layer_norm2.bias": "model.safetensors",
+        "vision_model.encoder.layers.11.layer_norm2.weight": "model.safetensors",
+        "vision_model.encoder.layers.11.mlp.fc1.bias": "model.safetensors",
+        "vision_model.encoder.layers.11.mlp.fc1.weight": "model.safetensors",
+        "vision_model.encoder.layers.11.mlp.fc2.bias": "model.safetensors",
+        "vision_model.encoder.layers.11.mlp.fc2.weight": "model.safetensors",
+        "vision_model.encoder.layers.11.self_attn.k_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.11.self_attn.k_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.11.self_attn.out_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.11.self_attn.out_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.11.self_attn.q_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.11.self_attn.q_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.11.self_attn.v_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.11.self_attn.v_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.2.layer_norm1.bias": "model.safetensors",
+        "vision_model.encoder.layers.2.layer_norm1.weight": "model.safetensors",
+        "vision_model.encoder.layers.2.layer_norm2.bias": "model.safetensors",
+        "vision_model.encoder.layers.2.layer_norm2.weight": "model.safetensors",
+        "vision_model.encoder.layers.2.mlp.fc1.bias": "model.safetensors",
+        "vision_model.encoder.layers.2.mlp.fc1.weight": "model.safetensors",
+        "vision_model.encoder.layers.2.mlp.fc2.bias": "model.safetensors",
+        "vision_model.encoder.layers.2.mlp.fc2.weight": "model.safetensors",
+        "vision_model.encoder.layers.2.self_attn.k_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.2.self_attn.k_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.2.self_attn.out_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.2.self_attn.out_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.2.self_attn.q_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.2.self_attn.q_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.2.self_attn.v_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.2.self_attn.v_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.3.layer_norm1.bias": "model.safetensors",
+        "vision_model.encoder.layers.3.layer_norm1.weight": "model.safetensors",
+        "vision_model.encoder.layers.3.layer_norm2.bias": "model.safetensors",
+        "vision_model.encoder.layers.3.layer_norm2.weight": "model.safetensors",
+        "vision_model.encoder.layers.3.mlp.fc1.bias": "model.safetensors",
+        "vision_model.encoder.layers.3.mlp.fc1.weight": "model.safetensors",
+        "vision_model.encoder.layers.3.mlp.fc2.bias": "model.safetensors",
+        "vision_model.encoder.layers.3.mlp.fc2.weight": "model.safetensors",
+        "vision_model.encoder.layers.3.self_attn.k_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.3.self_attn.k_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.3.self_attn.out_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.3.self_attn.out_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.3.self_attn.q_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.3.self_attn.q_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.3.self_attn.v_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.3.self_attn.v_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.4.layer_norm1.bias": "model.safetensors",
+        "vision_model.encoder.layers.4.layer_norm1.weight": "model.safetensors",
+        "vision_model.encoder.layers.4.layer_norm2.bias": "model.safetensors",
+        "vision_model.encoder.layers.4.layer_norm2.weight": "model.safetensors",
+        "vision_model.encoder.layers.4.mlp.fc1.bias": "model.safetensors",
+        "vision_model.encoder.layers.4.mlp.fc1.weight": "model.safetensors",
+        "vision_model.encoder.layers.4.mlp.fc2.bias": "model.safetensors",
+        "vision_model.encoder.layers.4.mlp.fc2.weight": "model.safetensors",
+        "vision_model.encoder.layers.4.self_attn.k_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.4.self_attn.k_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.4.self_attn.out_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.4.self_attn.out_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.4.self_attn.q_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.4.self_attn.q_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.4.self_attn.v_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.4.self_attn.v_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.5.layer_norm1.bias": "model.safetensors",
+        "vision_model.encoder.layers.5.layer_norm1.weight": "model.safetensors",
+        "vision_model.encoder.layers.5.layer_norm2.bias": "model.safetensors",
+        "vision_model.encoder.layers.5.layer_norm2.weight": "model.safetensors",
+        "vision_model.encoder.layers.5.mlp.fc1.bias": "model.safetensors",
+        "vision_model.encoder.layers.5.mlp.fc1.weight": "model.safetensors",
+        "vision_model.encoder.layers.5.mlp.fc2.bias": "model.safetensors",
+        "vision_model.encoder.layers.5.mlp.fc2.weight": "model.safetensors",
+        "vision_model.encoder.layers.5.self_attn.k_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.5.self_attn.k_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.5.self_attn.out_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.5.self_attn.out_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.5.self_attn.q_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.5.self_attn.q_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.5.self_attn.v_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.5.self_attn.v_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.6.layer_norm1.bias": "model.safetensors",
+        "vision_model.encoder.layers.6.layer_norm1.weight": "model.safetensors",
+        "vision_model.encoder.layers.6.layer_norm2.bias": "model.safetensors",
+        "vision_model.encoder.layers.6.layer_norm2.weight": "model.safetensors",
+        "vision_model.encoder.layers.6.mlp.fc1.bias": "model.safetensors",
+        "vision_model.encoder.layers.6.mlp.fc1.weight": "model.safetensors",
+        "vision_model.encoder.layers.6.mlp.fc2.bias": "model.safetensors",
+        "vision_model.encoder.layers.6.mlp.fc2.weight": "model.safetensors",
+        "vision_model.encoder.layers.6.self_attn.k_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.6.self_attn.k_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.6.self_attn.out_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.6.self_attn.out_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.6.self_attn.q_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.6.self_attn.q_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.6.self_attn.v_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.6.self_attn.v_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.7.layer_norm1.bias": "model.safetensors",
+        "vision_model.encoder.layers.7.layer_norm1.weight": "model.safetensors",
+        "vision_model.encoder.layers.7.layer_norm2.bias": "model.safetensors",
+        "vision_model.encoder.layers.7.layer_norm2.weight": "model.safetensors",
+        "vision_model.encoder.layers.7.mlp.fc1.bias": "model.safetensors",
+        "vision_model.encoder.layers.7.mlp.fc1.weight": "model.safetensors",
+        "vision_model.encoder.layers.7.mlp.fc2.bias": "model.safetensors",
+        "vision_model.encoder.layers.7.mlp.fc2.weight": "model.safetensors",
+        "vision_model.encoder.layers.7.self_attn.k_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.7.self_attn.k_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.7.self_attn.out_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.7.self_attn.out_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.7.self_attn.q_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.7.self_attn.q_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.7.self_attn.v_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.7.self_attn.v_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.8.layer_norm1.bias": "model.safetensors",
+        "vision_model.encoder.layers.8.layer_norm1.weight": "model.safetensors",
+        "vision_model.encoder.layers.8.layer_norm2.bias": "model.safetensors",
+        "vision_model.encoder.layers.8.layer_norm2.weight": "model.safetensors",
+        "vision_model.encoder.layers.8.mlp.fc1.bias": "model.safetensors",
+        "vision_model.encoder.layers.8.mlp.fc1.weight": "model.safetensors",
+        "vision_model.encoder.layers.8.mlp.fc2.bias": "model.safetensors",
+        "vision_model.encoder.layers.8.mlp.fc2.weight": "model.safetensors",
+        "vision_model.encoder.layers.8.self_attn.k_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.8.self_attn.k_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.8.self_attn.out_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.8.self_attn.out_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.8.self_attn.q_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.8.self_attn.q_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.8.self_attn.v_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.8.self_attn.v_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.9.layer_norm1.bias": "model.safetensors",
+        "vision_model.encoder.layers.9.layer_norm1.weight": "model.safetensors",
+        "vision_model.encoder.layers.9.layer_norm2.bias": "model.safetensors",
+        "vision_model.encoder.layers.9.layer_norm2.weight": "model.safetensors",
+        "vision_model.encoder.layers.9.mlp.fc1.bias": "model.safetensors",
+        "vision_model.encoder.layers.9.mlp.fc1.weight": "model.safetensors",
+        "vision_model.encoder.layers.9.mlp.fc2.bias": "model.safetensors",
+        "vision_model.encoder.layers.9.mlp.fc2.weight": "model.safetensors",
+        "vision_model.encoder.layers.9.self_attn.k_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.9.self_attn.k_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.9.self_attn.out_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.9.self_attn.out_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.9.self_attn.q_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.9.self_attn.q_proj.weight": "model.safetensors",
+        "vision_model.encoder.layers.9.self_attn.v_proj.bias": "model.safetensors",
+        "vision_model.encoder.layers.9.self_attn.v_proj.weight": "model.safetensors",
+        "vision_model.post_layernorm.bias": "model.safetensors",
+        "vision_model.post_layernorm.weight": "model.safetensors"
+    }
+}

preprocessor_config.json

@@ -0,0 +1,28 @@
+{
+  "do_convert_rgb": true,
+  "do_image_splitting": true,
+  "do_normalize": true,
+  "do_pad": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "Idefics3ImageProcessor",
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "max_image_size": {
+    "longest_edge": 512
+  },
+  "processor_class": "Idefics3Processor",
+  "resample": 1,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "longest_edge": 2048
+  }
+}

processor_config.json

@@ -0,0 +1,31 @@
+{
+  "image_processor": {
+    "do_convert_rgb": true,
+    "do_image_splitting": true,
+    "do_normalize": true,
+    "do_pad": true,
+    "do_rescale": true,
+    "do_resize": true,
+    "image_mean": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "image_processor_type": "Idefics3ImageProcessor",
+    "image_std": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "max_image_size": {
+      "longest_edge": 512
+    },
+    "resample": 1,
+    "rescale_factor": 0.00392156862745098,
+    "size": {
+      "longest_edge": 2048
+    }
+  },
+  "image_seq_len": 64,
+  "processor_class": "Idefics3Processor"
+}

special_tokens_map.json

@@ -0,0 +1,53 @@
+{
+  "additional_special_tokens": [
+    {
+      "content": "<fake_token_around_image>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false
+    },
+    {
+      "content": "<image>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false
+    },
+    {
+      "content": "<end_of_utterance>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false
+    }
+  ],
+  "bos_token": {
+    "content": "<|end_of_text|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<end_of_utterance>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|pad|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<|unk|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer_config.json

@@ -0,0 +1,29 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": "<|end_of_text|>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<end_of_utterance>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<fake_token_around_image>",
+    "<image>",
+    "<end_of_utterance>"
+  ],
+  "is_local": true,
+  "local_files_only": false,
+  "max_length": 8192,
+  "model_max_length": 8192,
+  "pad_to_multiple_of": null,
+  "pad_token": "<|pad|>",
+  "pad_token_type_id": 0,
+  "padding": "longest",
+  "padding_side": "left",
+  "processor_class": "Idefics3Processor",
+  "stride": 0,
+  "tokenizer_class": "GPT2Tokenizer",
+  "truncation": true,
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
+  "unk_token": "<|unk|>"
+}