Initial commit with existing Git LFS settings

.gitattributes

@@ -0,0 +1,38 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz 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
+*.onnx.data filter=lfs diff=lfs merge=lfs -text
+onnx/**/* filter=lfs diff=lfs merge=lfs -text
+**/.git* filter= diff= merge= text

.gitignore

@@ -0,0 +1 @@
+.DS_STORE

Makefile

@@ -0,0 +1,112 @@
+.SHELLFLAGS := -e -c
+SHELL := /bin/bash
+
+# Configuration variables
+NATIVE_ANDROID = ../Native-LLM-for-Android
+QWEN_VL_DIR = $(NATIVE_ANDROID)/Export_ONNX/QwenVL
+ONNX_SRC_DIR = $(QWEN_VL_DIR)/onnx
+ONNX_DEST_DIR = $(QWEN_VL_DIR)/onnx-dist
+STAGING_DIR = /tmp/transformers.js/staging
+TRANSFORMERS_JS_PATH = ../transformers.js
+ONNX_TOOLS_PATH = $(NATIVE_ANDROID)/ONNX_Tools
+
+# Python paths from venvs
+NATIVE_PYTHON = $(NATIVE_ANDROID)/.venv/bin/python3
+TRANSFORMERS_PYTHON = $(TRANSFORMERS_JS_PATH)/.venv/bin/python3
+
+# Model parts
+PARTS = A B C D E
+
+define progress_bar
+total=$$(echo $(1) | wc -w | tr -d ' '); \
+current=0; \
+for item in $(1); do \
+    current=$$((current + 1)); \
+    printf "\r   Progress: \033[1;32m["; \
+    for ((i=0; i<current*20/total; i++)); do printf "="; done; \
+    printf "\033[0m"; \
+    for ((i=current*20/total; i<20; i++)); do printf " "; done; \
+    printf "\033[1;32m]\033[0m $$current/$$total "; \
+    printf "\033[1;34m$$item\033[K\033[0m\n"; \
+    cmd="$(2)"; \
+    cmd=$$(echo "$$cmd" | sed "s|{}|$$item|g"); \
+    $$cmd; \
+done; \
+printf "\n"
+endef
+
+.PHONY: all all-in-one clean clean-large-files export fix-gpu-buffers quantize quantize-% slim
+
+all-in-one: export quantize clean-large-files slim fix-gpu-buffers
+	@echo "✨ All done! ONNX models exported, slimmed, quantized and fixed"
+
+export: export-abcd export-e
+	@echo "✅ Export complete"
+
+export-abcd:
+	@echo "🚀 Exporting parts A, B, C, D..."
+	cd ../Native-LLM-for-Android/Export_ONNX/QwenVL && \
+	../../.venv/bin/python3 QwenVL_Export_ABCD.py "Qwen/Qwen2-VL-2B-Instruct"
+
+export-e:
+	@echo "🚀 Exporting part E..."
+	cd ../Native-LLM-for-Android/Export_ONNX/QwenVL && \
+	../../.venv/bin/python3 QwenVL_Export_E.py "Qwen/Qwen2-VL-2B-Instruct"
+
+slim:
+	@echo "🗜️  Slimming ONNX models..."
+	@files=$$(find $(ONNX_SRC_DIR) -name "*.onnx" -type f ! -name "QwenVL_E.onnx"); \
+	$(call progress_bar,$$files,onnxslim --verbose {} {})
+	@echo "✅ Slimming complete"
+
+quantize:
+	@echo "⚡ Starting quantization..."
+	for part in $(PARTS); do \
+		$(MAKE) quantize-$$part || exit 1; \
+	done
+	@echo "✅ Quantization complete"
+
+quantize-%:
+	@echo "⚡ Quantizing part $*..."
+	mkdir -p $(ONNX_DEST_DIR)
+	cd $(TRANSFORMERS_JS_PATH) && \
+	mkdir -p $(STAGING_DIR) && \
+	rm -f $(STAGING_DIR)/* && \
+	ln -sf $$(realpath $(ONNX_SRC_DIR))/* $(STAGING_DIR)/ && \
+	find $(STAGING_DIR) -name "*_*_*.onnx_data" -delete && \
+	find $(STAGING_DIR) -name "*_*_*.onnx" -delete && \
+	find $(STAGING_DIR) -name "*.onnx" ! -name "QwenVL_$**.onnx" -delete && \
+	EXTRA_FLAGS=""; \
+	if [ "$*" = "A" ]; then EXTRA_FLAGS="--op_block_list Conv DynamicQuantizeLinear DequantizeLinear Resize"; fi; \
+	echo "Extra Flags for part $*: $$EXTRA_FLAGS" && \
+	PYTHONPATH=$(TRANSFORMERS_JS_PATH) .venv/bin/python3 -m scripts.quantize \
+		--input_folder '$(STAGING_DIR)' \
+		--output_folder '$(ONNX_DEST_DIR)' \
+		--mode q4f16 $$EXTRA_FLAGS
+
+clean-large-files:
+	@echo "🧹 Removing ONNX files over 2GB..."
+	cd $(ONNX_DEST_DIR) && \
+	for f in $$(find . -name "*.onnx" -type f); do \
+		total_size=0; \
+		if [ -f "$$f"".data" ]; then \
+			total_size=$$(( $$(stat -f %z "$$f") + $$(stat -f %z "$$f"".data") )); \
+		elif [ -f "$$f""_data" ]; then \
+			total_size=$$(( $$(stat -f %z "$$f") + $$(stat -f %z "$$f""_data") )); \
+		else \
+			total_size=$$(stat -f %z "$$f"); \
+		fi; \
+		size_mb=$$(( total_size / 1048576 )); \
+		if [ $$total_size -ge 2147483648 ]; then \
+			echo "   Removing $$f (size: $$size_mb MB)..."; \
+			rm -f "$$f" "$$f"".data" "$$f""_data"; \
+		fi \
+	done
+	@echo "✅ Large file cleanup complete"
+
+fix-gpu-buffers:
+	@echo "🔧 Fixing GPU buffers for E models..."
+	cd $(NATIVE_ANDROID) && \
+	files=$$(find $(ONNX_DEST_DIR) -name "QwenVL_E_*.onnx" -type f); \
+	$(call progress_bar,$$files, .venv/bin/python3 ONNX_Tools/clamp_for_gpu_buffers.py --overwrite {})
+	@echo "✅ GPU buffer fixes complete"

README.md

@@ -0,0 +1,5 @@
+---
+license: apache-2.0
+base_model:
+- Qwen/Qwen2-VL-2B-Instruct
+---

config.json

@@ -0,0 +1,52 @@
+{
+  "architectures": [
+    "Qwen2VLForConditionalGeneration"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "vision_start_token_id": 151652,
+  "vision_end_token_id": 151653,
+  "vision_token_id": 151654,
+  "image_token_id": 151655,
+  "video_token_id": 151656,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2_vl",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 2,
+  "rms_norm_eps": 1e-06,
+  "rope_theta": 1000000.0,
+  "sliding_window": 32768,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.41.2",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vision_config": {
+    "depth": 32,
+    "embed_dim": 1280,
+    "mlp_ratio": 4,
+    "num_heads": 16,
+    "in_chans": 3,
+    "hidden_size": 1536,
+    "patch_size": 14,
+    "spatial_merge_size": 2,
+    "spatial_patch_size": 14,
+    "temporal_patch_size": 2
+  },
+  "rope_scaling": {
+    "type": "mrope",
+    "mrope_section": [
+      16,
+      24,
+      24
+    ]
+  },
+  "vocab_size": 151936
+}

embeddings_bf16.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5387220a9d57908c2c6fa69bcfa64fda4234e59103fa74f56d07eaa6f9af2493
+size 466747392

generation_config.json

@@ -0,0 +1,15 @@
+{
+    "bos_token_id": 151643,
+    "pad_token_id": 151643,
+    "do_sample": true,
+    "eos_token_id": [
+      151645,
+      151643
+    ],
+    "repetition_penalty": 1.0,
+    "temperature": 0.01,
+    "top_p": 0.001,
+    "top_k": 1,
+    "transformers_version": "4.37.0"
+  }
+  

llm_config.json

@@ -0,0 +1,28 @@
+{
+    "hidden_size": 1536,
+    "layer_nums": 28,
+    "attention_mask": "float",
+    "key_value_shape": [
+        2,
+        1,
+        0,
+        2,
+        128
+    ],
+    "prompt_template": "<|im_start|>user\n%s<|im_end|>\n<|im_start|>assistant\n",
+    "is_visual": true,
+    "image_mean": [
+        122.7709383,
+        116.7460125,
+        104.09373615000001
+    ],
+    "image_norm": [
+        0.01459842661924292,
+        0.015007768493717056,
+        0.014220065717024088
+    ],
+    "image_size": 420,
+    "vision_start": 151652,
+    "vision_end": 151653,
+    "image_pad": 151655
+}

onnx/QwenVL_A_q4f16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdd6e4c85b5a835227106c01b31be8220eb4684026e726372a97c74cfdbcd983
+size 1330987067

onnx/QwenVL_B_q4f16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c5981ece4d144bc7f5352e56bb19d0d4b3bf22d1f8c472a106fcdcf83a9ebdf
+size 233983290

onnx/QwenVL_C_q4f16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc4f5f601f5ac0b16632e4dc953ce7009f8c2bf0c5e5c1553b5250cda832a68a
+size 6364

onnx/QwenVL_D_q4f16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5b46671d4d41a864d1390ef87eb5819e8c6fd044cded45e688301ae8eb7ab57
+size 25118

onnx/QwenVL_E_q4f16.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:599c81da0035adf98d77db9b5776e7070017887394d06dd901c4d72125f6fd2b
+size 996827324

tokenizer_config.json

@@ -0,0 +1,129 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": ["<|im_start|>", "<|im_end|>", "<|object_ref_start|>","<|object_ref_end|>","<|box_start|>","<|box_end|>","<|quad_start|>","<|quad_end|>","<|vision_start|>","<|vision_end|>","<|vision_pad|>","<|image_pad|>","<|video_pad|>"],
+  "bos_token": null,
+  "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "padding_side": "left",
+  "errors": "replace",
+  "model_max_length": 32768,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}