Upload folder using huggingface_hub

config.json

@@ -0,0 +1,316 @@
+{
+  "_name_or_path": "jellyho/TwinVLA-FT-50k",
+  "action_dim": 10,
+  "action_head": "DiT",
+  "action_len": 20,
+  "architectures": [
+    "Eagle2_1BTwinVLA"
+  ],
+  "denoiser": "FM",
+  "global_normalization": false,
+  "hz_interpolate": 20,
+  "interpolate_gripper": false,
+  "knowledge_insulation": false,
+  "model_path": null,
+  "model_type": "Eagle2_1BTwinVLA",
+  "modeling": "denoising",
+  "normalization": "quantile",
+  "num_readouts": 1,
+  "readout_token_as_eos": false,
+  "share_decoder": true,
+  "share_embed_tokens": true,
+  "share_vision": true,
+  "singlevla_config": {
+    "_attn_implementation_autoset": false,
+    "_attn_implementation_internal": null,
+    "_commit_hash": null,
+    "_name_or_path": "/home/jellyho/.cache/huggingface/hub/models--jellyho--TwinVLA-FT-50k/snapshots/85c86f3205c61063327aae4ef6dd5e6bce8fa093/singlevla_config/config.json",
+    "action_dim": 10,
+    "action_head": "DiT",
+    "action_head_hidden_dim": 1024,
+    "action_len": 20,
+    "add_cross_attention": false,
+    "aggregation": "false",
+    "architectures": [
+      "Eagle2_1BVLA"
+    ],
+    "auto_map": {},
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": null,
+    "chunk_size_feed_forward": 0,
+    "cross_attention_hidden_size": null,
+    "dataset_statistics_path": null,
+    "decoder_start_token_id": null,
+    "denoiser": "FM",
+    "diffusion_batch": 32,
+    "dit_size": "DiT-B",
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "downsample_ratio": 0.5,
+    "dynamic_image_size": true,
+    "early_stopping": false,
+    "efficient_loss": true,
+    "enable_cfg": true,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": null,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "force_image_size": 448,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "global_normalization": true,
+    "hz_interpolate": 20,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "image_size": 224,
+    "interpolate_gripper": false,
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "keep_aspect_ratio": false,
+    "knowledge_insulation": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "length_penalty": 1.0,
+    "llm_config": {
+      "_attn_implementation_autoset": true,
+      "_name_or_path": "./pretrained/Qwen2_5-0_5B-Instruct",
+      "add_cross_attention": false,
+      "architectures": [
+        "Qwen2ForCausalLM"
+      ],
+      "attention_dropout": 0.0,
+      "auto_map": {
+        "AutoConfig": "configuration_qwen2.Qwen2Config",
+        "AutoModel": "modeling_qwen2.Qwen2Model",
+        "AutoModelForCausalLM": "modeling_qwen2.Qwen2ForCausalLM"
+      },
+      "bad_words_ids": null,
+      "begin_suppress_tokens": null,
+      "bos_token_id": 151643,
+      "chunk_size_feed_forward": 0,
+      "cross_attention_hidden_size": null,
+      "decoder_start_token_id": null,
+      "diversity_penalty": 0.0,
+      "do_sample": false,
+      "early_stopping": false,
+      "encoder_no_repeat_ngram_size": 0,
+      "eos_token_id": 151645,
+      "exponential_decay_length_penalty": null,
+      "finetuning_task": null,
+      "forced_bos_token_id": null,
+      "forced_eos_token_id": null,
+      "hidden_act": "silu",
+      "hidden_size": 896,
+      "id2label": {
+        "0": "LABEL_0",
+        "1": "LABEL_1"
+      },
+      "initializer_range": 0.02,
+      "intermediate_size": 4864,
+      "is_decoder": false,
+      "is_encoder_decoder": false,
+      "label2id": {
+        "LABEL_0": 0,
+        "LABEL_1": 1
+      },
+      "length_penalty": 1.0,
+      "max_length": 20,
+      "max_position_embeddings": 32768,
+      "max_window_layers": 21,
+      "min_length": 0,
+      "model_type": "qwen2",
+      "no_repeat_ngram_size": 0,
+      "num_attention_heads": 14,
+      "num_beam_groups": 1,
+      "num_beams": 1,
+      "num_hidden_layers": 24,
+      "num_key_value_heads": 2,
+      "num_return_sequences": 1,
+      "output_attentions": false,
+      "output_hidden_states": false,
+      "output_scores": false,
+      "pad_token_id": null,
+      "prefix": null,
+      "problem_type": null,
+      "pruned_heads": {},
+      "remove_invalid_values": false,
+      "repetition_penalty": 1.0,
+      "return_dict": true,
+      "return_dict_in_generate": false,
+      "rms_norm_eps": 1e-06,
+      "rope_scaling": null,
+      "rope_theta": 1000000.0,
+      "sep_token_id": null,
+      "sliding_window": 32768,
+      "suppress_tokens": null,
+      "task_specific_params": null,
+      "temperature": 1.0,
+      "tf_legacy_loss": false,
+      "tie_encoder_decoder": false,
+      "tie_word_embeddings": true,
+      "tokenizer_class": null,
+      "top_k": 50,
+      "top_p": 1.0,
+      "torch_dtype": "bfloat16",
+      "torchscript": false,
+      "transformers_version": "4.50.0.dev0",
+      "typical_p": 1.0,
+      "use_bfloat16": false,
+      "use_cache": false,
+      "use_sliding_window": false,
+      "vocab_size": 151674
+    },
+    "loss_version": "v4",
+    "max_dynamic_patch": 12,
+    "max_length": 20,
+    "min_dynamic_patch": 1,
+    "min_length": 0,
+    "mlp_checkpoint": true,
+    "model_path": "nvidia/Eagle2-1B",
+    "model_type": "Eagle2_1BVLA",
+    "modeling": "denoising",
+    "no_repeat_ngram_size": 0,
+    "normalization": "quantile",
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_readouts": 1,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad2square": false,
+    "pad_token_id": null,
+    "pre_feature_reduction": false,
+    "prefix": null,
+    "problem_type": null,
+    "pruned_heads": {},
+    "ps_version": "v2",
+    "readout_token_as_eos": false,
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "return_text": null,
+    "select_layer": -1,
+    "sep_token_id": null,
+    "state_dim": 10,
+    "stopping_token": "|",
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "template": "qwen2-chat",
+    "test_denoising_steps": 10,
+    "tf_legacy_loss": false,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": true,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torch_dtype": "bfloat16",
+    "torchscript": false,
+    "train_denoising_steps": 100,
+    "typical_p": 1.0,
+    "use_backbone_lora": 0,
+    "use_bfloat16": false,
+    "use_llm_lora": 0,
+    "use_thumbnail": true,
+    "vision_config": {
+      "_attn_implementation_autoset": true,
+      "_name_or_path": "",
+      "add_cross_attention": false,
+      "architectures": [
+        "SiglipVisionModel"
+      ],
+      "attention_dropout": 0.0,
+      "auto_map": {
+        "AutoConfig": "configuration_siglip.SiglipVisionConfig",
+        "AutoModel": "modeling_siglip.SiglipVisionModel"
+      },
+      "bad_words_ids": null,
+      "begin_suppress_tokens": null,
+      "bos_token_id": null,
+      "chunk_size_feed_forward": 0,
+      "cross_attention_hidden_size": null,
+      "decoder_start_token_id": null,
+      "diversity_penalty": 0.0,
+      "do_sample": false,
+      "drop_path_rate": 0.1,
+      "early_stopping": false,
+      "encoder_no_repeat_ngram_size": 0,
+      "eos_token_id": null,
+      "exponential_decay_length_penalty": null,
+      "finetuning_task": null,
+      "forced_bos_token_id": null,
+      "forced_eos_token_id": null,
+      "hidden_act": "gelu_pytorch_tanh",
+      "hidden_size": 1152,
+      "id2label": {
+        "0": "LABEL_0",
+        "1": "LABEL_1"
+      },
+      "image_size": 448,
+      "intermediate_size": 4304,
+      "is_decoder": false,
+      "is_encoder_decoder": false,
+      "label2id": {
+        "LABEL_0": 0,
+        "LABEL_1": 1
+      },
+      "layer_norm_eps": 1e-06,
+      "length_penalty": 1.0,
+      "max_length": 20,
+      "min_length": 0,
+      "model_type": "siglip_vision_model",
+      "no_repeat_ngram_size": 0,
+      "num_attention_heads": 16,
+      "num_beam_groups": 1,
+      "num_beams": 1,
+      "num_channels": 3,
+      "num_hidden_layers": 27,
+      "num_image_tokens": 1024,
+      "num_return_sequences": 1,
+      "output_attentions": false,
+      "output_hidden_states": false,
+      "output_scores": false,
+      "pad_token_id": null,
+      "patch_size": 14,
+      "prefix": null,
+      "problem_type": null,
+      "projection_dim": 2048,
+      "projector_hidden_act": "gelu_fast",
+      "pruned_heads": {},
+      "remove_invalid_values": false,
+      "repetition_penalty": 1.0,
+      "return_dict": true,
+      "return_dict_in_generate": false,
+      "sep_token_id": null,
+      "suppress_tokens": null,
+      "task_specific_params": null,
+      "temperature": 1.0,
+      "tf_legacy_loss": false,
+      "tie_encoder_decoder": false,
+      "tie_word_embeddings": true,
+      "tokenizer_class": null,
+      "top_k": 50,
+      "top_p": 1.0,
+      "torch_dtype": "bfloat16",
+      "torchscript": false,
+      "transformers_version": "4.50.0.dev0",
+      "typical_p": 1.0,
+      "use_bfloat16": false,
+      "vision_use_head": false
+    },
+    "vocab_size": 151674,
+    "vocab_start": null
+  },
+  "singlevla_config_path": "/home/jellyho/.cache/huggingface/hub/models--jellyho--TwinVLA-FT-50k/snapshots/85c86f3205c61063327aae4ef6dd5e6bce8fa093/singlevla_config/config.json",
+  "singlevla_pretrained_path": null,
+  "state_dim": 10,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.50.0.dev0"
+}

dataset_statistics.json

@@ -0,0 +1,318 @@
+{
+  "aloha_handover_box": {
+    "action": {
+      "mean": [
+        0.322445809841156,
+        -0.07136808335781097,
+        0.16350804269313812,
+        0.6155836582183838,
+        0.025826361030340195,
+        -0.6291796565055847,
+        -0.12873482704162598,
+        0.9684553742408752,
+        -0.05324767157435417,
+        0.32048356533050537,
+        0.3553526699542999,
+        -0.017269087955355644,
+        0.25354716181755066,
+        0.9467635154724121,
+        -0.10934817045927048,
+        -0.16364239156246185,
+        0.10808151960372925,
+        0.9652227759361267,
+        -0.08225584030151367,
+        0.6809535622596741
+      ],
+      "std": [
+        0.07454723119735718,
+        0.08869021385908127,
+        0.07996607571840286,
+        0.33460575342178345,
+        0.19985812902450562,
+        0.26947152614593506,
+        0.12514586746692657,
+        0.030899910256266594,
+        0.16146336495876312,
+        0.9473028779029846,
+        0.06487792730331421,
+        0.03894990682601929,
+        0.027652453631162643,
+        0.10490529984235764,
+        0.1838434785604477,
+        0.14178059995174408,
+        0.2033252716064453,
+        0.06656672805547714,
+        0.064219169318676,
+        0.732288122177124
+      ],
+      "max": [
+        0.48683926463127136,
+        0.0484432689845562,
+        0.31490612030029297,
+        0.99891197681427,
+        0.4277522563934326,
+        0.06322141736745834,
+        0.4004654884338379,
+        0.9999857544898987,
+        0.3100079298019409,
+        1.0,
+        0.5334027409553528,
+        0.08494444936513901,
+        0.36568865180015564,
+        0.9999882578849792,
+        0.2546274662017822,
+        0.1172015443444252,
+        0.7982608079910278,
+        0.9999992251396179,
+        0.20094169676303864,
+        1.0
+      ],
+      "min": [
+        0.1422317922115326,
+        -0.2763901352882385,
+        -0.0600760243833065,
+        -0.14848311245441437,
+        -0.6282482743263245,
+        -0.9999129176139832,
+        -0.42181891202926636,
+        0.7404066324234009,
+        -0.6676974296569824,
+        -1.0,
+        0.1786160171031952,
+        -0.1845615804195404,
+        0.1687021553516388,
+        0.2762398421764374,
+        -0.7479667067527771,
+        -0.8485982418060303,
+        -0.2597721517086029,
+        0.6015138626098633,
+        -0.3933228552341461,
+        -1.0
+      ],
+      "q01": [
+        0.1950138956308365,
+        -0.24691226959228516,
+        -0.015285035967826844,
+        -0.04555398792028427,
+        -0.4452396559715271,
+        -0.996303243637085,
+        -0.3760478734970093,
+        0.8516808867454528,
+        -0.46342918753623963,
+        -1.0,
+        0.21926841557025908,
+        -0.1317625629901886,
+        0.1978745412826538,
+        0.5117229986190795,
+        -0.6376786828041077,
+        -0.6609986042976379,
+        -0.19099083304405212,
+        0.6930621123313904,
+        -0.2356126993894577,
+        -1.0
+      ],
+      "q99": [
+        0.47150796771049497,
+        0.038070930540561675,
+        0.28182336688041676,
+        0.9817836880683899,
+        0.3871919810771942,
+        -0.1345064049959186,
+        0.20285944879054985,
+        0.9992118668556214,
+        0.2293877118825912,
+        1.0,
+        0.49810330152511595,
+        0.0599309906363487,
+        0.3309180569648742,
+        0.9995350050926208,
+        0.1829529863595952,
+        0.03216676786541939,
+        0.7132800936698909,
+        0.9997488117218017,
+        0.08941484957933345,
+        1.0
+      ],
+      "mask": [
+        true,
+        true,
+        true,
+        false,
+        false,
+        false,
+        false,
+        false,
+        false,
+        false,
+        true,
+        true,
+        true,
+        false,
+        false,
+        false,
+        false,
+        false,
+        false,
+        false
+      ]
+    },
+    "proprio": {
+      "mean": [
+        0.3199884295463562,
+        -0.06322039663791656,
+        0.15364451706409454,
+        0.6062000393867493,
+        0.06611454486846924,
+        -0.6415385603904724,
+        -0.09341777116060257,
+        0.9627954959869385,
+        0.031210733577609062,
+        0.3975388705730438,
+        0.3515152037143707,
+        -0.018268277868628502,
+        0.2389151155948639,
+        0.9404945969581604,
+        -0.11243173480033875,
+        -0.20223592221736908,
+        0.10952325910329819,
+        0.9571213722229004,
+        -0.07971788197755814,
+        0.5135871171951294
+      ],
+      "std": [
+        0.07216393202543259,
+        0.07979436963796616,
+        0.08181631565093994,
+        0.3255034387111664,
+        0.15539786219596863,
+        0.2940728962421417,
+        0.1470193713903427,
+        0.0702429711818695,
+        0.19176805019378662,
+        0.34666576981544495,
+        0.06400732696056366,
+        0.03745417669415474,
+        0.032076891511678696,
+        0.10509590804576874,
+        0.17774498462677002,
+        0.1389302909374237,
+        0.20042844116687775,
+        0.08065077662467957,
+        0.13744565844535828,
+        0.2912951111793518
+      ],
+      "max": [
+        0.4756927788257599,
+        0.08933446556329727,
+        0.374162495136261,
+        0.9999960064888,
+        0.6316872835159302,
+        0.4776941239833832,
+        0.5646718740463257,
+        1.0,
+        0.9942742586135864,
+        0.9009114503860474,
+        0.525163471698761,
+        0.07913362979888916,
+        0.41776642203330994,
+        0.9999624490737915,
+        0.22881703078746796,
+        0.25566166639328003,
+        0.7931073904037476,
+        1.0,
+        0.8460071086883545,
+        0.7628055810928345
+      ],
+      "min": [
+        0.15654076635837555,
+        -0.24968452751636505,
+        -0.005626573693007231,
+        -0.12249734997749329,
+        -0.7090452313423157,
+        -1.0,
+        -0.8474201560020447,
+        -0.10166153311729431,
+        -0.7077121138572693,
+        -0.8731642365455627,
+        0.18484462797641754,
+        -0.16924390196800232,
+        0.1543029397726059,
+        0.2589649260044098,
+        -0.7204840183258057,
+        -0.8606683015823364,
+        -0.24437052011489868,
+        0.35231029987335205,
+        -0.9356772899627686,
+        -0.8731642365455627
+      ],
+      "q01": [
+        0.19023896217346192,
+        -0.22257453262805937,
+        -0.0035876664519309998,
+        -0.01533471692353487,
+        -0.2564712131023407,
+        -0.9997995805740356,
+        -0.35406238079071045,
+        0.6182524561882019,
+        -0.3273082673549652,
+        -0.11072122812271118,
+        0.2202826726436615,
+        -0.12777820765972137,
+        0.17903109550476073,
+        0.48722490668296814,
+        -0.6251857328414917,
+        -0.6860081076622009,
+        -0.17821017920970916,
+        0.6367788648605347,
+        -0.562646164894104,
+        -0.22518292188644407
+      ],
+      "q99": [
+        0.4643557775020599,
+        0.0528511093556879,
+        0.3235322630405426,
+        0.9950041770935059,
+        0.42506823778152464,
+        0.17489826560020202,
+        0.36311201095581025,
+        0.9999364614486694,
+        0.7507713937759389,
+        0.7889349985122672,
+        0.49394992589950554,
+        0.051381032466888375,
+        0.33517178773879985,
+        0.999138171672821,
+        0.16684073746204348,
+        0.055357125997543326,
+        0.7065780282020567,
+        0.999832546710968,
+        0.513761196136471,
+        0.6879761123657224
+      ],
+      "mask": [
+        true,
+        true,
+        true,
+        false,
+        false,
+        false,
+        false,
+        false,
+        false,
+        false,
+        true,
+        true,
+        true,
+        false,
+        false,
+        false,
+        false,
+        false,
+        false,
+        false
+      ]
+    },
+    "num_transitions": 11829,
+    "num_trajectories": 50
+  }
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec89f703d0b817e56b17386fc63e774c105c317c3d45a93552d54bc348468407
+size 2889539864

singlevla_config/config.json

@@ -0,0 +1,230 @@
+{
+  "_commit_hash": null,
+  "_name_or_path": "/home/jellyho/.cache/huggingface/hub/models--jellyho--TwinVLA-FT-50k/snapshots/85c86f3205c61063327aae4ef6dd5e6bce8fa093/singlevla_config/config.json",
+  "action_dim": 10,
+  "action_head": "DiT",
+  "action_head_hidden_dim": 1024,
+  "action_len": 20,
+  "aggregation": "false",
+  "architectures": [
+    "Eagle2_1BVLA"
+  ],
+  "auto_map": {},
+  "dataset_statistics_path": null,
+  "denoiser": "FM",
+  "diffusion_batch": 32,
+  "dit_size": "DiT-B",
+  "downsample_ratio": 0.5,
+  "dynamic_image_size": true,
+  "efficient_loss": true,
+  "enable_cfg": true,
+  "force_image_size": 448,
+  "global_normalization": true,
+  "hz_interpolate": 20,
+  "image_size": 224,
+  "interpolate_gripper": false,
+  "keep_aspect_ratio": false,
+  "knowledge_insulation": false,
+  "llm_config": {
+    "_attn_implementation_autoset": true,
+    "_name_or_path": "./pretrained/Qwen2_5-0_5B-Instruct",
+    "add_cross_attention": false,
+    "architectures": [
+      "Qwen2ForCausalLM"
+    ],
+    "attention_dropout": 0.0,
+    "auto_map": {
+      "AutoConfig": "configuration_qwen2.Qwen2Config",
+      "AutoModel": "modeling_qwen2.Qwen2Model",
+      "AutoModelForCausalLM": "modeling_qwen2.Qwen2ForCausalLM"
+    },
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": 151643,
+    "chunk_size_feed_forward": 0,
+    "cross_attention_hidden_size": null,
+    "decoder_start_token_id": null,
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "early_stopping": false,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": 151645,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "hidden_act": "silu",
+    "hidden_size": 896,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "initializer_range": 0.02,
+    "intermediate_size": 4864,
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "max_position_embeddings": 32768,
+    "max_window_layers": 21,
+    "min_length": 0,
+    "model_type": "qwen2",
+    "no_repeat_ngram_size": 0,
+    "num_attention_heads": 14,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_hidden_layers": 24,
+    "num_key_value_heads": 2,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": null,
+    "prefix": null,
+    "problem_type": null,
+    "pruned_heads": {},
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "rms_norm_eps": 1e-06,
+    "rope_scaling": null,
+    "rope_theta": 1000000.0,
+    "sep_token_id": null,
+    "sliding_window": 32768,
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "tf_legacy_loss": false,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": true,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torch_dtype": "bfloat16",
+    "torchscript": false,
+    "transformers_version": "4.50.0.dev0",
+    "typical_p": 1.0,
+    "use_bfloat16": false,
+    "use_cache": false,
+    "use_sliding_window": false,
+    "vocab_size": 151674
+  },
+  "loss_version": "v4",
+  "max_dynamic_patch": 12,
+  "min_dynamic_patch": 1,
+  "mlp_checkpoint": true,
+  "model_path": "nvidia/Eagle2-1B",
+  "model_type": "Eagle2_1BVLA",
+  "modeling": "denoising",
+  "normalization": "quantile",
+  "num_readouts": 1,
+  "pad2square": false,
+  "pre_feature_reduction": false,
+  "ps_version": "v2",
+  "readout_token_as_eos": false,
+  "return_text": null,
+  "select_layer": -1,
+  "state_dim": 10,
+  "stopping_token": "|",
+  "template": "qwen2-chat",
+  "test_denoising_steps": 10,
+  "torch_dtype": "bfloat16",
+  "train_denoising_steps": 100,
+  "transformers_version": null,
+  "use_backbone_lora": 0,
+  "use_llm_lora": 0,
+  "use_thumbnail": true,
+  "vision_config": {
+    "_attn_implementation_autoset": true,
+    "_name_or_path": "",
+    "add_cross_attention": false,
+    "architectures": [
+      "SiglipVisionModel"
+    ],
+    "attention_dropout": 0.0,
+    "auto_map": {
+      "AutoConfig": "configuration_siglip.SiglipVisionConfig",
+      "AutoModel": "modeling_siglip.SiglipVisionModel"
+    },
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": null,
+    "chunk_size_feed_forward": 0,
+    "cross_attention_hidden_size": null,
+    "decoder_start_token_id": null,
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "drop_path_rate": 0.1,
+    "early_stopping": false,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": null,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 1152,
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "image_size": 448,
+    "intermediate_size": 4304,
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "layer_norm_eps": 1e-06,
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "min_length": 0,
+    "model_type": "siglip_vision_model",
+    "no_repeat_ngram_size": 0,
+    "num_attention_heads": 16,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_channels": 3,
+    "num_hidden_layers": 27,
+    "num_image_tokens": 1024,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": null,
+    "patch_size": 14,
+    "prefix": null,
+    "problem_type": null,
+    "projection_dim": 2048,
+    "projector_hidden_act": "gelu_fast",
+    "pruned_heads": {},
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "sep_token_id": null,
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "tf_legacy_loss": false,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": true,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torch_dtype": "bfloat16",
+    "torchscript": false,
+    "transformers_version": "4.50.0.dev0",
+    "typical_p": 1.0,
+    "use_bfloat16": false,
+    "vision_use_head": false
+  },
+  "vocab_size": 151674,
+  "vocab_start": null
+}