Upload TED KD model (oob)

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* 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
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,72 @@
+---
+license: apache-2.0
+base_model: Qwen/Qwen3-1.7B
+tags:
+- safety
+- classifier
+- knowledge-distillation
+- ted
+- dia-guard
+language:
+- en
+---
+
+# Shield-Qwen3-1.7B-KD-TED-Qwen3-4B-SafeRL-OOB
+
+**Student:** `Qwen/Qwen3-1.7B`
+**Teacher:** `Qwen/Qwen3-4B-SafeRL`
+**KD method:** TED
+**Scenario:** OOB (out-of-box — neither teacher nor student was fine-tuned on DIA-GUARD before KD)
+
+Part of the **DIA-GUARD** dialect-aware safety classifier suite. This checkpoint
+is the result of distilling an off-the-shelf 4B/8B safety teacher into a smaller
+student using 50K dialect-stratified samples from the DIA-GUARD train split, and
+evaluating on the full 181,874-sample dialect holdout test.
+
+## Test Set Results
+
+| Metric | Value |
+|--------|-------|
+| **Accuracy** | **0.5758** |
+| **Precision** | 0.6299 |
+| **Recall** | 0.5980 |
+| **F1** | 0.5567 |
+| Test samples | 181,874 |
+
+### Per-class breakdown
+
+| Class | Precision | Recall | F1 | Support |
+|-------|-----------|--------|-----|---------|
+| safe | 0.5219 | 0.8570 | 0.6487 | 83,140 |
+| unsafe | 0.7379 | 0.3391 | 0.4646 | 98,734 |
+
+### Confusion matrix
+
+|               | Predicted safe | Predicted unsafe |
+|---------------|---------------:|-----------------:|
+| **Actual safe**   | TN=71,247 | FP=11,893 |
+| **Actual unsafe** | FN=65,258 | TP=33,476 |
+
+## Training Setup
+
+| | |
+|---|---|
+| Method | TED |
+| Teacher | Qwen/Qwen3-4B-SafeRL |
+| Student base | Qwen/Qwen3-1.7B |
+| Train data | 50,000 dialect-stratified DIA-GUARD samples |
+| Epochs | 1 |
+| Framework | HuggingFace transformers + accelerate |
+
+## How to use
+
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+tokenizer = AutoTokenizer.from_pretrained("jsl5710/Shield-Qwen3-1.7B-KD-TED-Qwen3-4B-SafeRL-OOB")
+model = AutoModelForCausalLM.from_pretrained("jsl5710/Shield-Qwen3-1.7B-KD-TED-Qwen3-4B-SafeRL-OOB", torch_dtype="bfloat16")
+
+prompt = "Classify the following text as safe or unsafe.\n\nText: Hello\n\nAnswer:"
+inputs = tokenizer(prompt, return_tensors="pt")
+out = model.generate(**inputs, max_new_tokens=8)
+print(tokenizer.decode(out[0], skip_special_tokens=True))
+```

alignment_metadata.json

@@ -0,0 +1,121 @@
+{
+  "student_num_layers": 28,
+  "teacher_num_layers": 36,
+  "student_hidden": 2048,
+  "teacher_hidden": 2560,
+  "alignment": [
+    [
+      0,
+      0
+    ],
+    [
+      1,
+      1
+    ],
+    [
+      2,
+      3
+    ],
+    [
+      3,
+      4
+    ],
+    [
+      4,
+      5
+    ],
+    [
+      5,
+      6
+    ],
+    [
+      6,
+      8
+    ],
+    [
+      7,
+      9
+    ],
+    [
+      8,
+      10
+    ],
+    [
+      9,
+      12
+    ],
+    [
+      10,
+      13
+    ],
+    [
+      11,
+      14
+    ],
+    [
+      12,
+      16
+    ],
+    [
+      13,
+      17
+    ],
+    [
+      14,
+      18
+    ],
+    [
+      15,
+      19
+    ],
+    [
+      16,
+      21
+    ],
+    [
+      17,
+      22
+    ],
+    [
+      18,
+      23
+    ],
+    [
+      19,
+      25
+    ],
+    [
+      20,
+      26
+    ],
+    [
+      21,
+      27
+    ],
+    [
+      22,
+      29
+    ],
+    [
+      23,
+      30
+    ],
+    [
+      24,
+      31
+    ],
+    [
+      25,
+      32
+    ],
+    [
+      26,
+      34
+    ],
+    [
+      27,
+      35
+    ]
+  ],
+  "num_align_layers": 28
+}

chat_template.jinja

@@ -0,0 +1,89 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

config.json

@@ -0,0 +1,64 @@
+{
+  "architectures": [
+    "Qwen3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "dtype": "bfloat16",
+  "eos_token_id": 151645,
+  "head_dim": 128,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 6144,
+  "layer_types": [
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 40960,
+  "max_window_layers": 28,
+  "model_type": "qwen3",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 8,
+  "output_hidden_states": true,
+  "pad_token_id": null,
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "rope_theta": 1000000,
+    "rope_type": "default"
+  },
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.5.0",
+  "use_cache": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}

generation_config.json

@@ -0,0 +1,13 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "temperature": 0.6,
+  "top_k": 20,
+  "top_p": 0.95,
+  "transformers_version": "5.5.0"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78c1ac2e3a0cc5002a9cfbde2604cf0b8a7dd180d8fd30b1fd8658c3b7b2a20d
+size 3441185608

task_aware_filters.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:12b17e3ae9bbfe2ab40c0553b6ab8039917698621844cd6ae4806c02b4d9bb07
+size 293611973

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
+size 11422650

tokenizer_config.json

@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_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|>"
+  ],
+  "is_local": false,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}