Upload folder using huggingface_hub

.gitattributes

@@ -34,3 +34,11 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 tokenizer.json filter=lfs diff=lfs merge=lfs -text
+onnx/audio_encoder.onnx_data filter=lfs diff=lfs merge=lfs -text
+onnx/audio_encoder_fp16.onnx_data filter=lfs diff=lfs merge=lfs -text
+onnx/decoder_model_merged.onnx_data filter=lfs diff=lfs merge=lfs -text
+onnx/decoder_model_merged_fp16.onnx_data filter=lfs diff=lfs merge=lfs -text
+onnx/embed_tokens.onnx_data filter=lfs diff=lfs merge=lfs -text
+onnx/embed_tokens_fp16.onnx_data filter=lfs diff=lfs merge=lfs -text
+onnx/vision_encoder.onnx_data filter=lfs diff=lfs merge=lfs -text
+onnx/vision_encoder_fp16.onnx_data filter=lfs diff=lfs merge=lfs -text

config.json

@@ -12,6 +12,7 @@
     "conf_num_hidden_layers": 2,
     "conf_reduction_factor": 4,
     "conf_residual_weight": 0.5,
+    "dtype": "bfloat16",
     "gradient_clipping": 10000000000.0,
     "hidden_size": 64,
     "input_feat_size": 128,
@@ -42,7 +43,6 @@
         2
       ]
     ],
-    "torch_dtype": "bfloat16",
     "vocab_offset": 262272,
     "vocab_size": 128
   },
@@ -50,6 +50,7 @@
   "audio_token_id": 262273,
   "boa_token_id": 256000,
   "boi_token_id": 255999,
+  "dtype": "bfloat16",
   "eoa_token_id": 262272,
   "eoi_token_id": 262144,
   "eos_token_id": [
@@ -72,6 +73,9 @@
     "altup_num_inputs": 4,
     "attention_bias": false,
     "attention_dropout": 0.0,
+    "bos_token_id": 2,
+    "dtype": "bfloat16",
+    "eos_token_id": 1,
     "final_logit_softcapping": 30.0,
     "head_dim": 32,
     "hidden_activation": "gelu_pytorch_tanh",
@@ -97,21 +101,30 @@
     "num_hidden_layers": 4,
     "num_key_value_heads": 1,
     "num_kv_shared_layers": 2,
+    "pad_token_id": 0,
     "rms_norm_eps": 1e-06,
-    "rope_local_base_freq": 10000.0,
-    "rope_scaling": null,
-    "rope_theta": 1000000.0,
+    "rope_parameters": {
+      "full_attention": {
+        "rope_theta": 1000000.0,
+        "rope_type": "default"
+      },
+      "sliding_attention": {
+        "rope_theta": 10000.0,
+        "rope_type": "default"
+      }
+    },
     "sliding_window": 512,
-    "torch_dtype": "bfloat16",
+    "tie_word_embeddings": true,
     "use_cache": true,
     "vocab_size": 262400,
     "vocab_size_per_layer_input": 262144
   },
-  "torch_dtype": "bfloat16",
-  "transformers_version": "4.53.2",
+  "tie_word_embeddings": true,
+  "transformers_version": "5.3.0.dev0",
   "vision_config": {
     "architecture": "mobilenetv5_300m_enc",
     "do_pooling": false,
+    "dtype": "bfloat16",
     "hidden_size": 2048,
     "initializer_range": 0.02,
     "label_names": [
@@ -277,9 +290,24 @@
     "model_type": "gemma3n_vision",
     "num_classes": 2,
     "rms_norm_eps": 1e-06,
-    "torch_dtype": "bfloat16",
     "vocab_offset": 262144,
     "vocab_size": 128
   },
-  "vision_soft_tokens_per_image": 256
-}
+  "vision_soft_tokens_per_image": 256,
+  "transformers.js_config": {
+    "use_external_data_format": {
+      "decoder_model_merged.onnx": 1,
+      "embed_tokens.onnx": 1,
+      "audio_encoder.onnx": 1,
+      "vision_encoder.onnx": 1,
+      "decoder_model_merged_fp16.onnx": 1,
+      "embed_tokens_fp16.onnx": 1,
+      "audio_encoder_fp16.onnx": 1,
+      "vision_encoder_fp16.onnx": 1
+    },
+    "kv_cache_dtype": {
+      "q4f16": "float16",
+      "fp16": "float16"
+    }
+  }
+}

generation_config.json

@@ -9,6 +9,6 @@
   "pad_token_id": 0,
   "top_k": 64,
   "top_p": 0.95,
-  "transformers_version": "4.53.2",
-  "trust_remote_code": true
+  "transformers_version": "5.3.0.dev0",
+  "trust_remote_code": false
 }

onnx/audio_encoder.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99d0c9ae5f59d629428e1b7b95679a99608fca45e3d4db5738f71f2b38e7d8c8
+size 85829

onnx/audio_encoder.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad01ced648acb6d8be96d0b496090af96a7cf5e6ff6738ab7eb5f5b7993cc36e
+size 1185280

onnx/audio_encoder_fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c3eef215db75dcca84489c384939b45371765ef42c221f9389d03837f907c96
+size 88445

onnx/audio_encoder_fp16.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:46bb776b1ddc24ffdda01b65c604ffa6818a36970b394d4cac2ceb46535384c2
+size 591360

onnx/decoder_model_merged.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23dad85daa5513a2fec1de07a5f29735beeeac6a987a191813147c4eb98f8e4f
+size 221367

onnx/decoder_model_merged.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:582a0f3703579ab00ba5c5d53be8433ec1e0a9e961500768f774d261288e7072
+size 42157056

onnx/decoder_model_merged_fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9753fd50a89b0da45648b5546e3a38fa7ff28befda15b01062500eb34aad665f
+size 244912

onnx/decoder_model_merged_fp16.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1efc35ba3fe0f28d52485f4ea064e6f1acd0833bd3dbe044614a3b06f5df24c9
+size 21073920

onnx/embed_tokens.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:841e9b1cd993484b3eaf7fa6ecd70710cbfb3bc7826f15daab502fd42d22c268
+size 2518

onnx/embed_tokens.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3914f667ab7d56d508a3a7185c7dc58cb24eaf62989c4013e35b120813ca562b
+size 41975808

onnx/embed_tokens_fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe288814cb8a3f8f28a3b05825921087ffbd7e603b7bccff47aa4d117fbe8fc2
+size 2987

onnx/embed_tokens_fp16.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a40d3d603e327ded01e83138127ed2426666da0a3e4dd83d55922562838a982f
+size 20987904

onnx/vision_encoder.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77de514a12de8b120542a24ca52e303a4fe881cc18c6d49586421d79c6b64d13
+size 98161

onnx/vision_encoder.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:138633e802dbe842c925eafacacdcf459d14ed241169fbcb1ccbeeaeb6ce0e83
+size 2586496

onnx/vision_encoder_fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bbff8e53a81de7ef0cc8c477f760061ec78cedfbe0868a8f0292c221c6db81c1
+size 96373

onnx/vision_encoder_fp16.onnx_data

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20d1d561e54868f967dd6c626cffb1fdbe9f2b1fda177a760f62bf2074510890
+size 1294528

processor_config.json

@@ -1,5 +1,97 @@
 {
   "audio_seq_length": 188,
+  "feature_extractor": {
+    "crop_size": null,
+    "data_format": "channels_first",
+    "default_to_square": false,
+    "device": null,
+    "disable_grouping": null,
+    "dither": 0.0,
+    "do_center_crop": null,
+    "do_convert_rgb": null,
+    "do_normalize": false,
+    "do_rescale": true,
+    "do_resize": true,
+    "feature_extractor_type": "Gemma3nAudioFeatureExtractor",
+    "feature_size": 128,
+    "fft_length": 1024,
+    "fft_overdrive": true,
+    "frame_length": 512,
+    "hop_length": 160,
+    "image_mean": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "image_processor_type": "SiglipImageProcessorFast",
+    "image_seq_length": 256,
+    "image_std": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "input_data_format": null,
+    "input_scale_factor": 1.0,
+    "max_frequency": 7600.0,
+    "mel_floor": 1e-05,
+    "min_frequency": 125.0,
+    "padding_side": "right",
+    "padding_value": 0.0,
+    "per_bin_mean": null,
+    "per_bin_stddev": null,
+    "preemphasis": 0.97,
+    "preemphasis_htk_flavor": true,
+    "resample": 2,
+    "rescale_factor": 0.00392156862745098,
+    "return_attention_mask": true,
+    "return_tensors": null,
+    "sampling_rate": 16000,
+    "size": {
+      "height": 768,
+      "width": 768
+    }
+  },
+  "image_processor": {
+    "data_format": "channels_first",
+    "default_to_square": false,
+    "dither": 0.0,
+    "do_normalize": false,
+    "do_rescale": true,
+    "do_resize": true,
+    "feature_size": 128,
+    "fft_length": 1024,
+    "fft_overdrive": true,
+    "frame_length": 512,
+    "hop_length": 160,
+    "image_mean": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "image_processor_type": "SiglipImageProcessorFast",
+    "image_seq_length": 256,
+    "image_std": [
+      0.5,
+      0.5,
+      0.5
+    ],
+    "input_scale_factor": 1.0,
+    "max_frequency": 7600.0,
+    "mel_floor": 1e-05,
+    "min_frequency": 125.0,
+    "padding_side": "right",
+    "padding_value": 0.0,
+    "preemphasis": 0.97,
+    "preemphasis_htk_flavor": true,
+    "resample": 2,
+    "rescale_factor": 0.00392156862745098,
+    "return_attention_mask": true,
+    "sampling_rate": 16000,
+    "size": {
+      "height": 768,
+      "width": 768
+    }
+  },
   "image_seq_length": 256,
   "processor_class": "Gemma3nProcessor"
 }

tokenizer.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b6c35ee648c07754b44cd9e371c75d4caa05c4504910b7ad29b1847ee9d8ba5d
-size 33442553
+oid sha256:44cb3d7d545cf895311e004d9a2b2ce823be5eb84c9aa31f73858b607c44c924
+size 20366294