Upload folder using huggingface_hub

README.md

@@ -1,3 +1,11 @@
----
-license: mit
----
+# Janus-Pro-7B WebGPU
+
+WebGPU-optimized Janus-Pro-7B for transformers.js.
+
+## Usage
+```javascript
+import { loadJanus } from './usage_example.js';
+const { model, processor } = await loadJanus();
+```
+
+Ready for browser deployment! 🚀

config.json

@@ -0,0 +1,66 @@
+{
+  "aligner_config": {
+    "cls": "MlpProjector",
+    "model_type": "aligner",
+    "params": {
+      "depth": 2,
+      "input_dim": 1024,
+      "n_embed": 4096,
+      "projector_type": "mlp_gelu"
+    }
+  },
+  "gen_aligner_config": {
+    "cls": "MlpProjector",
+    "model_type": "gen_aligner",
+    "params": {
+      "depth": 2,
+      "input_dim": 8,
+      "n_embed": 4096,
+      "projector_type": "mlp_gelu"
+    }
+  },
+  "gen_head_config": {
+    "cls": "vision_head",
+    "model_type": "gen_head",
+    "params": {
+      "image_token_embed": 4096,
+      "image_token_size": 16384,
+      "n_embed": 4096
+    }
+  },
+  "gen_vision_config": {
+    "cls": "VQ-16",
+    "model_type": "gen_vision",
+    "params": {
+      "image_token_size": 16384,
+      "n_embed": 8
+    }
+  },
+  "language_config": {
+    "max_position_embeddings": 16384,
+    "model_type": "llama",
+    "num_hidden_layers": 30,
+    "torch_dtype": "bfloat16",
+    "vocab_size": 102400
+  },
+  "model_type": "janus",
+  "torch_dtype": "float16",
+  "transformers_version": "4.33.1",
+  "vision_config": {
+    "cls": "CLIPVisionTower",
+    "model_type": "vision",
+    "params": {
+      "image_size": 384,
+      "model_name": "siglip_large_patch16_384",
+      "select_feature": "same",
+      "select_layer": -1
+    }
+  },
+  "architectures": [
+    "JanusForConditionalGeneration"
+  ],
+  "use_cache": true,
+  "webgpu_compatible": true,
+  "quantization": "q4f16",
+  "pipeline_tag": "text-to-image"
+}

generation_config.json

@@ -0,0 +1,12 @@
+{
+  "bos_token_id": 100000,
+  "eos_token_id": 100001,
+  "pad_token_id": 151643,
+  "max_length": 2048,
+  "max_new_tokens": 1024,
+  "do_sample": true,
+  "temperature": 0.7,
+  "top_p": 0.9,
+  "use_cache": true,
+  "num_image_tokens": 576
+}

model_index.json

@@ -0,0 +1,20 @@
+{
+  "model_name": "Janus-Pro-7B",
+  "model_type": "multimodal",
+  "architecture": "janus",
+  "format": "onnx",
+  "quantization": "q4f16",
+  "device": "webgpu",
+  "total_size_mb": 4935.2,
+  "components_count": 6,
+  "status": "complete",
+  "capabilities": [
+    "text-to-image",
+    "image-to-text",
+    "multimodal-chat"
+  ],
+  "transformers_js_compatible": true,
+  "webgpu_optimized": true,
+  "export_date": "2025-09-27",
+  "notes": "All 6 components successfully exported for WebGPU deployment"
+}

onnx/gen_head_fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22eb336ab33e4d0b9ee5c09c2d3906682c4aeff408a189a2eb45077ff3f1de30
+size 335628308

onnx/gen_img_embeds_fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:184a1b1eb70e501a8a5831e7ab5193364f973a4d9946dc12caf410cf85294f13
+size 67125624

onnx/language_model_q4f16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:df9b2ef1d20b7314c4c877b75939bbbed5dacb3aa1f4e75fc47fe0b6e88e0e3a
+size 1073941695

onnx/lm_head_fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6599c2b692b6efd30d7b2ab4ae04434ba352f8f4ed4fdc967ca71ef69debdf8f
+size 1677721863

onnx/prepare_inputs_embeds_q4.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b94e5d5ccaf0a02a804dbcb35c7f0d0b57f22fd30b1f187c2ec797fb34df6c5b
+size 1677721872

onnx/vision_encoder_fp16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c17e3ed46cb81ac30d8cf1c9c6c723afc77108a52a63df4c6b74e7c1dd8431c
+size 1215274800

preprocessor_config.json

@@ -0,0 +1,23 @@
+{
+  "background_color": [
+    127,
+    127,
+    127
+  ],
+  "do_normalize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "VLMImageProcessor",
+  "image_size": 384,
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "min_size": 14,
+  "processor_class": "VLChatProcessor",
+  "rescale_factor": 0.00392156862745098
+}

processor_config.json

@@ -0,0 +1,9 @@
+{
+  "add_special_token": false,
+  "ignore_id": -100,
+  "image_tag": "<image_placeholder>",
+  "mask_prompt": true,
+  "num_image_tokens": 576,
+  "processor_class": "VLChatProcessor",
+  "sft_format": "deepseek"
+}

special_tokens_map.json

@@ -0,0 +1,16 @@
+{
+  "additional_special_tokens": [
+    "<image_placeholder>",
+    "<patch_placeholder>",
+    "<|ref|>",
+    "<|/ref|>",
+    "<|det|>",
+    "<|/det|>",
+    "<|grounding|>",
+    "<|User|>",
+    "<|Assistant|>"
+  ],
+  "bos_token": "<｜begin▁of▁sentence｜>",
+  "eos_token": "<｜end▁of▁sentence｜>",
+  "pad_token": "<｜▁pad▁｜>"
+}

tokenizer_config.json

@@ -0,0 +1,10 @@
+{
+  "bos_token": "<｜begin▁of▁sentence｜>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<｜end▁of▁sentence｜>",
+  "model_max_length": 16384,
+  "pad_token": null,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": null,
+  "use_default_system_prompt": true
+}

usage_example.js

@@ -0,0 +1,13 @@
+// Janus-Pro-7B WebGPU Usage
+import { AutoProcessor, AutoModelForCausalLM } from "@huggingface/transformers";
+
+async function loadJanus() {
+    const processor = await AutoProcessor.from_pretrained("./janus-pro-7b-webgpu");
+    const model = await AutoModelForCausalLM.from_pretrained("./janus-pro-7b-webgpu", {
+        device: "webgpu",
+        dtype: "q4f16"
+    });
+    return { model, processor };
+}
+
+export { loadJanus };