Upload instruction_following model with 21 experts (14.3B params)

.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 +1,328 @@
+---
+license: apache-2.0
+datasets:
+  - AmanPriyanshu/GPT-OSS-20B-MoE-expert-activations
+language:
+  - en
+pipeline_tag: text-generation
+tags:
+  - mixture-of-experts
+  - moe
+  - expert-pruning
+  - gpt-oss
+  - openai
+  - reasoning
+  - instruction-following
+  - specialized
+  - efficient
+  - transformer
+  - causal-lm
+  - text-generation
+  - pytorch
+  - pruned-model
+  - domain-specific
+---
+
+# Instruction Following GPT-OSS Model (21 Experts)
+
+**Project**: https://amanpriyanshu.github.io/GPT-OSS-MoE-ExpertFingerprinting/
+
+<div align="center">
+
+### 👥 Follow the Authors
+
+**Aman Priyanshu**
+[![LinkedIn](https://img.shields.io/badge/LinkedIn-0077B5?style=for-the-badge&logo=linkedin&logoColor=white)](https://www.linkedin.com/in/aman-priyanshu/)
+[![Twitter](https://img.shields.io/badge/Twitter-1DA1F2?style=for-the-badge&logo=twitter&logoColor=white)](https://x.com/AmanPriyanshu6)
+[![Website](https://img.shields.io/badge/Website-FF7139?style=for-the-badge&logo=firefox&logoColor=white)](https://amanpriyanshu.github.io/)
+
+**Supriti Vijay**
+[![LinkedIn](https://img.shields.io/badge/LinkedIn-0077B5?style=for-the-badge&logo=linkedin&logoColor=white)](https://www.linkedin.com/in/supriti-vijay/)
+[![Twitter](https://img.shields.io/badge/Twitter-1DA1F2?style=for-the-badge&logo=twitter&logoColor=white)](https://x.com/SupritiVijay)
+[![Website](https://img.shields.io/badge/Website-FF7139?style=for-the-badge&logo=firefox&logoColor=white)](https://supritivijay.github.io/)
+
+</div>
+
+## Introduction
+
+This is a pruned variant of OpenAI's GPT-OSS-20B model, reduced to 21 experts per layer based on activation patterns from the [AmanPriyanshu/GPT-OSS-20B MoE Expert Activations dataset](https://huggingface.co/datasets/AmanPriyanshu/GPT-OSS-20B-MoE-expert-activations). We analyzed router decisions across evaluation benchmarks to identify and retain experts most relevant for instruction following tasks.
+
+**⚠️ Experimental Model**: This is an experimental pruned model that may not work well - check the [examples below](#model-examples) to see if the outputs meet your needs before use.
+
+This pruning approach reduces the model size while attempting to preserve performance on the target domain.
+
+## Model Architecture & Statistics
+
+| Metric | Value |
+|--------|-------|
+| **Base Model** | openai/gpt-oss-20b |
+| **Architecture** | Mixture-of-Experts Transformer |
+| **Total Parameters** | ~14.3B (pruned from 21B) |
+| **Original Experts per Layer** | 32 |
+| **Pruned Experts per Layer** | 21 |
+| **Layers** | 24 |
+| **Top-k Routing** | 4 |
+| **Context Length** | 128K tokens |
+| **Attention Heads** | 64 (Query), 8 (Key-Value) |
+| **Residual Dimension** | 2880 |
+| **Attention Pattern** | Alternating dense & sliding window (128 tokens) |
+| **Positional Encoding** | RoPE (Rotary Position Embedding) |
+| **Normalization** | RMSNorm |
+| **Precision** | BF16 |
+| **License** | Apache 2.0 |
+| **Specialization** | Instruction Following |
+
+## Pruning Methodology
+
+### What is Expert Pruning?
+Mixture-of-Experts models contain multiple specialized sub-networks (experts) per layer. During inference, only a subset of experts are activated for each token. Expert pruning involves:
+
+1. **Analyzing Usage Patterns**: Tracking which experts activate most frequently for specific tasks
+2. **Removing Underutilized Experts**: Discarding experts with low activation rates for the target domain
+3. **Preserving Router Functionality**: Maintaining the routing mechanism with fewer available experts
+
+### Our Approach
+- **Data-Driven Selection**: Used activation patterns from instruction following evaluation tasks
+- **Systematic Reduction**: Reduced from 32 to 21 experts per layer
+- **No Retraining**: Direct removal without additional training steps
+
+## Performance & Applications
+
+### Pruning Benefits
+- **Smaller Memory Footprint**: 65.6% of original expert parameters
+- **Reduced Computational Load**: Fewer routing decisions during inference
+- **Focused Capabilities**: Retains experts relevant to instruction following tasks
+
+### Use Cases
+- **Speculative Decoding**: Draft model for full GPT-OSS-20B
+- **Resource-Constrained Deployment**: Edge devices, mobile applications
+- **Research**: Study expert specialization in MoE models
+- **Fine-tuning**: Smaller base model for domain adaptation
+
+*Note: Performance may vary depending on how well the pruned experts match your specific use case.*
+
+## Motivation & Expert Selection
+
+This instruction-following model leverages experts that excelled at constraint satisfaction tasks from Tulu3 Persona Instruction Following dataset. These experts specialize in precise adherence to user specifications and formatting requirements.
+
+The expert selection process utilized our comprehensive analysis of router activation patterns across multiple evaluation benchmarks:
+
+- **GPQA**: Graduate-level questions in physics, chemistry, biology (Diamond & Expert subsets)
+- **MMLU/MMLU-Pro**: Comprehensive knowledge across 57+ subjects including science, medicine, law
+- **SORRY-Bench**: Safety evaluation across harmful content categories  
+- **Tulu3**: Persona-driven instruction following with verifiable constraints
+- **Polyglot-or-Not**: Multilingual factual completion tasks
+
+By identifying experts that consistently activated for instruction following tasks, we created this specialized model that maintains domain expertise while significantly reducing computational requirements from 32 to 21 experts per layer.
+
+## Dataset & Analysis Foundation
+
+This model is based on analysis from the **GPT-OSS-20B MoE Expert Activations dataset** available at:
+🔗 **https://huggingface.co/datasets/AmanPriyanshu/GPT-OSS-20B-MoE-expert-activations**
+
+The dataset contains router activation patterns from OpenAI's GPT-OSS-20B model across diverse evaluation benchmarks, enabling the creation of these domain-optimized models through systematic expert pruning.
+
+### Pruning Methodology
+Our approach involves:
+1. **Activation Analysis**: Comprehensive evaluation of expert usage patterns across domain-specific tasks
+2. **Expert Ranking**: Identification of the most frequently activated experts for target domains  
+3. **Systematic Pruning**: Reduction from 32 to 21 experts while preserving router functionality
+4. **Quality Validation**: Testing to ensure maintained performance on target tasks
+
+*This is a direct pruning approach - no additional training was performed. The model inherits all capabilities from the original GPT-OSS-20B with focused expert selection.*
+
+## Usage
+
+### CPU Inference
+
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+
+# Load the specialized model on CPU
+model = AutoModelForCausalLM.from_pretrained(
+    "AmanPriyanshu/gpt-oss-14.3b-specialized-instruction_following-pruned-moe-only-21-experts", 
+    torch_dtype=torch.bfloat16, 
+    device_map="cpu", 
+    trust_remote_code=True
+)
+tokenizer = AutoTokenizer.from_pretrained("AmanPriyanshu/gpt-oss-14.3b-specialized-instruction_following-pruned-moe-only-21-experts")
+
+# Generate with the model
+messages = [
+    {"role": "user", "content": "Write a formal email to a professor requesting a meeting, including: subject line, greeting, purpose, proposed times, and professional closing."}
+]
+
+inputs = tokenizer.apply_chat_template(
+    messages, 
+    add_generation_prompt=True, 
+    return_tensors="pt", 
+    return_dict=True,
+    reasoning_effort="medium"
+)
+
+# Ensure inputs are on the same device as model
+inputs = {k: v.to(model.device) for k, v in inputs.items()}
+
+outputs = model.generate(
+    **inputs, 
+    max_new_tokens=512,
+    do_sample=True,
+    temperature=0.1,
+    top_p=0.9,
+    pad_token_id=tokenizer.eos_token_id,
+    eos_token_id=tokenizer.eos_token_id
+)
+
+# Decode only the generated part
+input_length = inputs['input_ids'].shape[1]
+response_tokens = outputs[0][input_length:]
+response = tokenizer.decode(response_tokens, skip_special_tokens=True)
+print(response)
+```
+
+### Apple Silicon (MPS) Inference
+
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
+
+# Check MPS availability and load model
+device = "mps" if torch.backends.mps.is_available() else "cpu"
+
+model = AutoModelForCausalLM.from_pretrained(
+    "AmanPriyanshu/gpt-oss-14.3b-specialized-instruction_following-pruned-moe-only-21-experts", 
+    torch_dtype=torch.float16,  # Better MPS compatibility
+    device_map=device, 
+    trust_remote_code=True,
+    low_cpu_mem_usage=True
+)
+tokenizer = AutoTokenizer.from_pretrained("AmanPriyanshu/gpt-oss-14.3b-specialized-instruction_following-pruned-moe-only-21-experts")
+
+# Generate with the model
+messages = [
+    {"role": "user", "content": "Write a formal email to a professor requesting a meeting, including: subject line, greeting, purpose, proposed times, and professional closing."}
+]
+
+inputs = tokenizer.apply_chat_template(
+    messages, 
+    add_generation_prompt=True, 
+    return_tensors="pt", 
+    return_dict=True,
+    reasoning_effort="medium"
+)
+
+# Move inputs to model device
+inputs = {k: v.to(model.device) if isinstance(v, torch.Tensor) else v for k, v in inputs.items()}
+
+# Use torch.no_grad for MPS stability
+with torch.no_grad():
+    outputs = model.generate(
+        **inputs, 
+        max_new_tokens=512,
+        do_sample=True,
+        temperature=0.1,
+        top_p=0.9,
+        pad_token_id=tokenizer.eos_token_id,
+        eos_token_id=tokenizer.eos_token_id,
+        use_cache=True
+    )
+
+# Decode only the generated part
+input_length = inputs['input_ids'].shape[1]
+response_tokens = outputs[0][input_length:]
+response = tokenizer.decode(response_tokens, skip_special_tokens=True)
+print(response)
+```
+
+### GPU Inference
+
+For CUDA GPU inference, the models work with minimal modifications from the CPU example above. Simply change:
+```python
+device_map="auto"  # Will automatically use GPU if available
+torch_dtype=torch.bfloat16  # or torch.float16
+```
+
+## Model Examples
+
+The following examples demonstrate the model's performance on instruction following tasks:
+
+### Example 1
+**Question:** Write a formal email to a professor requesting a meeting, including: subject line, greeting, purpose, proposed times, and professional closing.
+
+**Response:**
+```
+<|channel|>analysis<|message|>Need formal email.<|end|><|start|>assistant<|channel|>final<|message|>**Subject:** Request for Meeting – Discussion of Research Project Progress  
+
+Dear Professor [Last Name],
+
+I hope this message finds you well. I am writing to request a brief meeting with you to discuss the current status of my research project on [brief topic], as well as to seek your guidance on the next steps.
+
+I would be available at the following times over the next week:
+
+- Tuesday, [Date], 10:00 – 11:00 a.m.  
+- Wednesday, [Date], 2:00 – 3:00 p.m.  
+- Friday, [Date], 9:30 – 10:30 a.m.  
+
+If any of these slots are convenient for you, please let me know, and I will adjust my schedule accordingly. If you prefer a different time, I am happy to accommodate.
+
+Thank you for your time and consideration. I look forward to your guidance.
+
+Sincerely,
+
+[Your Full Name]  
+[Your Program / Department]  
+[University Name]  
+[Phone Number] | [Email Address]<|return|>
+```
+
+---
+
+### Example 2
+**Question:** Create a numbered list of steps for making scrambled eggs, using exactly 6 steps.
+
+**Response:**
+```
+<|channel|>analysis<|message|>Need 6 steps numbered.<|end|><|start|>assistant<|channel|>final<|message|>1. **Heat the pan** – Place a non‑stick skillet over medium heat and let it warm up for a minute or two.  
+2. **Add a little fat** – Drop a teaspoon of butter, oil, or a pat of softened butter into the pan and swirl it around until it melts and coats the surface.  
+3. **Crack the eggs** – Crack the desired number of eggs into a bowl, then whisk them together with a splash of milk or water (about 1 tablespoon per egg) and a pinch of salt or pepper.  
+4. **Pour the mixture** – Gently pour the whisked eggs into the hot, gre‑ed pan, letting them spread out evenly.  
+5. **Cook and stir** – Let the eggs sit for 30–60 seconds, then use a spatula to stir gently, lifting the edges and folding the center until the whites are set but the yolks remain slightly runny.  
+6. **Serve immediately** – Once the eggs are cooked to your liking, slide them onto a plate, add any extra seasonings or toppings, and enjoy while hot.<|return|>
+```
+
+---
+
+### Example 3
+**Question:** Summarize the water cycle in exactly 50 words, no more, no less.
+
+**Response:**
+```
+<|channel|>analysis<|message|>Need 50 words exactly. Let's craft: "Water evaporates from oceans, lakes, and soil, forming clouds. Precipitation returns water to Earth as rain, snow, or sleet. Streams and rivers transport it to larger bodies, where it infiltrates groundwater. Plants absorb it, releasing vapor through transpiration. The cycle repeats, sustaining ecosystems and human life." Count words. Let's count: Water(1) evaporates2 from3 oceans,4 lakes,5 and6 soil,7 forming8 clouds.9 Precipitation10 returns11 water12 to13 Earth14 as15 rain,16 snow,17 or18 sleet.19 Streams20 and21 rivers22 transport23 it24 to25 larger26 bodies,27 where28 it29 infiltrates30 groundwater.31 Plants32 absorb33 it,34 releasing35 vapor36 through37 transpiration.38 The39 cycle40 repeats,41 sustaining42 ecosystems43 and44 human45 life46. Need 4 more words. Add "for"47 "all"48 "generations"49 "daily"50. Final sentence: "The cycle repeats, sustaining ecosystems and human life for all generations daily." Count again: The1 cycle2 repeats,3 sustaining4 ecosystems5 and6 human7 life8 for9 all10 generations11 daily12. That's 12 words, too many. Need 4 words total. We already had 46 words. Need 4 more words to reach 50. Add "for"47 "all"48 "generations"49 "daily"50. But we need to attach to sentence. Could say: "The cycle repeats, sustaining ecosystems and human life for all generations daily." Count words: The1 cycle2 repeats,3 sustaining4 ecosystems5 and6 human7 life8 for9 all10 generations11 daily12. That's 12 words, too many. We need only 4 words added, not 12. So we need to add 4 words somewhere else, not whole phrase. Maybe add "in" "every" "season" "year". Let's add at end: "in every season year". Count words: in1 every2 season3 year4. That's 4 words. Add to end of paragraph. So final paragraph ends with "...daily in every season year." Count again entire paragraph. Let's recount all words.
+
+Water1 evaporates2 from3 oceans,4 lakes,5 and6 soil,7 forming8 clouds.9 Precipitation10 returns11 water12 to13
+```
+
+---
+
+## Citation
+
+If you use this model in your research, please cite:
+
+```bibtex
+@misc{priyanshu2025gptoss,
+  title={{GPT-OSS MoE Expert Fingerprinting: Analyzing Expert Activation Patterns in Mixture of Experts Models}},
+  author={Priyanshu, Aman and Vijay, Supriti},
+  year={2025},
+  howpublished={\url{https://amanpriyanshu.github.io/GPT-OSS-MoE-ExpertFingerprinting/}},
+  note={Interactive analysis tool for expert activation patterns in MoE architectures}
+}
+```
+
+## References & Resources
+
+- **Original Model**: [OpenAI GPT-OSS Model Card](https://openai.com/index/introducing-gpt-oss/)
+- **Model Hub**: [GPT-OSS-20B on Hugging Face](https://huggingface.co/openai/gpt-oss-20b)
+- **Expert Analysis Dataset**: [GPT-OSS-20B MoE Expert Activations](https://huggingface.co/datasets/AmanPriyanshu/GPT-OSS-20B-MoE-expert-activations)
+- **Project Page**: [GPT-OSS MoE Expert Fingerprinting](https://amanpriyanshu.github.io/GPT-OSS-MoE-ExpertFingerprinting/)
+- **GitHub Repository**: [OpenAI GPT-OSS](https://github.com/openai/gpt-oss)

chat_template.jinja

@@ -0,0 +1,331 @@
+{#-
+  In addition to the normal inputs of `messages` and `tools`, this template also accepts the
+  following kwargs:
+  - "builtin_tools": A list, can contain "browser" and/or "python".
+  - "model_identity": A string that optionally describes the model identity.
+  - "reasoning_effort": A string that describes the reasoning effort, defaults to "medium".
+ #}
+
+{#- Tool Definition Rendering ============================================== #}
+{%- macro render_typescript_type(param_spec, required_params, is_nullable=false) -%}
+    {%- if param_spec.type == "array" -%}
+        {%- if param_spec['items'] -%}
+            {%- if param_spec['items']['type'] == "string" -%}
+                {{- "string[]" }}
+            {%- elif param_spec['items']['type'] == "number" -%}
+                {{- "number[]" }}
+            {%- elif param_spec['items']['type'] == "integer" -%}
+                {{- "number[]" }}
+            {%- elif param_spec['items']['type'] == "boolean" -%}
+                {{- "boolean[]" }}
+            {%- else -%}
+                {%- set inner_type = render_typescript_type(param_spec['items'], required_params) -%}
+                {%- if inner_type == "object | object" or inner_type|length > 50 -%}
+                    {{- "any[]" }}
+                {%- else -%}
+                    {{- inner_type + "[]" }}
+                {%- endif -%}
+            {%- endif -%}
+            {%- if param_spec.nullable -%}
+                {{- " | null" }}
+            {%- endif -%}
+        {%- else -%}
+            {{- "any[]" }}
+            {%- if param_spec.nullable -%}
+                {{- " | null" }}
+            {%- endif -%}
+        {%- endif -%}
+    {%- elif param_spec.type is defined and param_spec.type is iterable and param_spec.type is not string and param_spec.type is not mapping and param_spec.type[0] is defined -%}
+        {#- Handle array of types like ["object", "object"] from Union[dict, list] #}
+        {%- if param_spec.type | length > 1 -%}
+            {{- param_spec.type | join(" | ") }}
+        {%- else -%}
+            {{- param_spec.type[0] }}
+        {%- endif -%}
+    {%- elif param_spec.oneOf -%}
+        {#- Handle oneOf schemas - check for complex unions and fallback to any #}
+        {%- set has_object_variants = false -%}
+        {%- for variant in param_spec.oneOf -%}
+            {%- if variant.type == "object" -%}
+                {%- set has_object_variants = true -%}
+            {%- endif -%}
+        {%- endfor -%}
+        {%- if has_object_variants and param_spec.oneOf|length > 1 -%}
+            {{- "any" }}
+        {%- else -%}
+            {%- for variant in param_spec.oneOf -%}
+                {{- render_typescript_type(variant, required_params) -}}
+                {%- if variant.description %}
+                    {{- "// " + variant.description }}
+                {%- endif -%}
+                {%- if variant.default is defined %}
+                    {{ "// default: " + variant.default|tojson }}
+                {%- endif -%}
+                {%- if not loop.last %}
+                    {{- " | " }}
+                {% endif -%}
+            {%- endfor -%}
+        {%- endif -%}
+    {%- elif param_spec.type == "string" -%}
+        {%- if param_spec.enum -%}
+            {{- '"' + param_spec.enum|join('" | "') + '"' -}}
+        {%- else -%}
+            {{- "string" }}
+            {%- if param_spec.nullable %}
+                {{- " | null" }}
+            {%- endif -%}
+        {%- endif -%}
+    {%- elif param_spec.type == "number" -%}
+        {{- "number" }}
+    {%- elif param_spec.type == "integer" -%}
+        {{- "number" }}
+    {%- elif param_spec.type == "boolean" -%}
+        {{- "boolean" }}
+
+    {%- elif param_spec.type == "object" -%}
+        {%- if param_spec.properties -%}
+            {{- "{\n" }}
+            {%- for prop_name, prop_spec in param_spec.properties.items() -%}
+                {{- prop_name -}}
+                {%- if prop_name not in (param_spec.required or []) -%}
+                    {{- "?" }}
+                {%- endif -%}
+                {{- ": " }}
+                {{ render_typescript_type(prop_spec, param_spec.required or []) }}
+                {%- if not loop.last -%}
+                    {{-", " }}
+                {%- endif -%}
+            {%- endfor -%}
+            {{- "}" }}
+        {%- else -%}
+            {{- "object" }}
+        {%- endif -%}
+    {%- else -%}
+        {{- "any" }}
+    {%- endif -%}
+{%- endmacro -%}
+
+{%- macro render_tool_namespace(namespace_name, tools) -%}
+    {{- "## " + namespace_name + "\n\n" }}
+    {{- "namespace " + namespace_name + " {\n\n" }}
+    {%- for tool in tools %}
+        {%- set tool = tool.function %}
+        {{- "// " + tool.description + "\n" }}
+        {{- "type "+ tool.name + " = " }}
+        {%- if tool.parameters and tool.parameters.properties %}
+            {{- "(_: {\n" }}
+            {%- for param_name, param_spec in tool.parameters.properties.items() %}
+                {%- if param_spec.description %}
+                    {{- "// " + param_spec.description + "\n" }}
+                {%- endif %}
+                {{- param_name }}
+                {%- if param_name not in (tool.parameters.required or []) -%}
+                    {{- "?" }}
+                {%- endif -%}
+                {{- ": " }}
+                {{- render_typescript_type(param_spec, tool.parameters.required or []) }}
+                {%- if param_spec.default is defined -%}
+                    {%- if param_spec.enum %}
+                        {{- ", // default: " + param_spec.default }}
+                    {%- elif param_spec.oneOf %}
+                        {{- "// default: " + param_spec.default }}
+                    {%- else %}
+                        {{- ", // default: " + param_spec.default|tojson }}
+                    {%- endif -%}
+                {%- endif -%}
+                {%- if not loop.last %}
+                    {{- ",\n" }}
+                {%- else %}
+                    {{- ",\n" }}
+                {%- endif -%}
+            {%- endfor %}
+            {{- "}) => any;\n\n" }}
+        {%- else -%}
+            {{- "() => any;\n\n" }}
+        {%- endif -%}
+    {%- endfor %}
+    {{- "} // namespace " + namespace_name }}
+{%- endmacro -%}
+
+{%- macro render_builtin_tools(browser_tool, python_tool) -%}
+    {%- if browser_tool %}
+        {{- "## browser\n\n" }}
+        {{- "// Tool for browsing.\n" }}
+        {{- "// The `cursor` appears in brackets before each browsing display: `[{cursor}]`.\n" }}
+        {{- "// Cite information from the tool using the following format:\n" }}
+        {{- "// `【{cursor}†L{line_start}(-L{line_end})?】`, for example: `【6†L9-L11】` or `【8†L3】`.\n" }}
+        {{- "// Do not quote more than 10 words directly from the tool output.\n" }}
+        {{- "// sources=web (default: web)\n" }}
+        {{- "namespace browser {\n\n" }}
+        {{- "// Searches for information related to `query` and displays `topn` results.\n" }}
+        {{- "type search = (_: {\n" }}
+        {{- "query: string,\n" }}
+        {{- "topn?: number, // default: 10\n" }}
+        {{- "source?: string,\n" }}
+        {{- "}) => any;\n\n" }}
+        {{- "// Opens the link `id` from the page indicated by `cursor` starting at line number `loc`, showing `num_lines` lines.\n" }}
+        {{- "// Valid link ids are displayed with the formatting: `【{id}†.*】`.\n" }}
+        {{- "// If `cursor` is not provided, the most recent page is implied.\n" }}
+        {{- "// If `id` is a string, it is treated as a fully qualified URL associated with `source`.\n" }}
+        {{- "// If `loc` is not provided, the viewport will be positioned at the beginning of the document or centered on the most relevant passage, if available.\n" }}
+        {{- "// Use this function without `id` to scroll to a new location of an opened page.\n" }}
+        {{- "type open = (_: {\n" }}
+        {{- "id?: number | string, // default: -1\n" }}
+        {{- "cursor?: number, // default: -1\n" }}
+        {{- "loc?: number, // default: -1\n" }}
+        {{- "num_lines?: number, // default: -1\n" }}
+        {{- "view_source?: boolean, // default: false\n" }}
+        {{- "source?: string,\n" }}
+        {{- "}) => any;\n\n" }}
+        {{- "// Finds exact matches of `pattern` in the current page, or the page given by `cursor`.\n" }}
+        {{- "type find = (_: {\n" }}
+        {{- "pattern: string,\n" }}
+        {{- "cursor?: number, // default: -1\n" }}
+        {{- "}) => any;\n\n" }}
+        {{- "} // namespace browser\n\n" }}
+    {%- endif -%}
+
+    {%- if python_tool %}
+        {{- "## python\n\n" }}
+        {{- "Use this tool to execute Python code in your chain of thought. The code will not be shown to the user. This tool should be used for internal reasoning, but not for code that is intended to be visible to the user (e.g. when creating plots, tables, or files).\n\n" }}
+        {{- "When you send a message containing Python code to python, it will be executed in a stateful Jupyter notebook environment. python will respond with the output of the execution or time out after 120.0 seconds. The drive at '/mnt/data' can be used to save and persist user files. Internet access for this session is UNKNOWN. Depends on the cluster.\n\n" }}
+    {%- endif -%}
+{%- endmacro -%}
+
+{#- System Message Construction ============================================ #}
+{%- macro build_system_message() -%}
+    {%- if model_identity is not defined %}
+        {%- set model_identity = "You are ChatGPT, a large language model trained by OpenAI." %}
+    {%- endif %}
+    {{- model_identity + "\n" }}
+    {{- "Knowledge cutoff: 2024-06\n" }}
+    {{- "Current date: " + strftime_now("%Y-%m-%d") + "\n\n" }}
+    {%- if reasoning_effort is not defined %}
+        {%- set reasoning_effort = "medium" %}
+    {%- endif %}
+    {{- "Reasoning: " + reasoning_effort + "\n\n" }}
+    {%- if builtin_tools %}
+        {{- "# Tools\n\n" }}
+        {%- set available_builtin_tools = namespace(browser=false, python=false) %}
+        {%- for tool in builtin_tools %}
+            {%- if tool == "browser" %}
+                {%- set available_builtin_tools.browser = true %}
+            {%- elif tool == "python" %}
+                {%- set available_builtin_tools.python = true %}
+            {%- endif %}
+        {%- endfor %}
+        {{- render_builtin_tools(available_builtin_tools.browser, available_builtin_tools.python) }}
+    {%- endif -%}
+    {{- "# Valid channels: analysis, commentary, final. Channel must be included for every message." }}
+    {%- if tools -%}
+        {{- "\nCalls to these tools must go to the commentary channel: 'functions'." }}
+    {%- endif -%}
+{%- endmacro -%}
+
+{#- Main Template Logic ================================================= #}
+{#- Set defaults #}
+
+{#- Render system message #}
+{{- "<|start|>system<|message|>" }}
+{{- build_system_message() }}
+{{- "<|end|>" }}
+
+{#- Extract developer message #}
+{%- if messages[0].role == "developer" or messages[0].role == "system" %}
+    {%- set developer_message = messages[0].content %}
+    {%- set loop_messages = messages[1:] %}
+{%- else %}
+    {%- set developer_message = "" %}
+    {%- set loop_messages = messages %}
+{%- endif %}
+
+{#- Render developer message #}
+{%- if developer_message or tools %}
+    {{- "<|start|>developer<|message|>" }}
+    {%- if developer_message %}
+        {{- "# Instructions\n\n" }}
+        {{- developer_message }}
+        {{- "\n\n" }}
+    {%- endif %}
+    {%- if tools -%}
+        {{- "# Tools\n\n" }}
+        {{- render_tool_namespace("functions", tools) }}
+    {%- endif -%}
+    {{- "<|end|>" }}
+{%- endif %}
+
+{#- Render messages #}
+{%- set last_tool_call = namespace(name=none) %}
+{%- for message in loop_messages -%}
+    {#- At this point only assistant/user/tool messages should remain #}
+    {%- if message.role == 'assistant' -%}
+        {#- Checks to ensure the messages are being passed in the format we expect #}
+        {%- if "content" in message %}
+            {%- if "<|channel|>analysis<|message|>" in message.content or "<|channel|>final<|message|>" in message.content %}
+                {{- raise_exception("You have passed a message containing <|channel|> tags in the content field. Instead of doing this, you should pass analysis messages (the string between '<|message|>' and '<|end|>') in the 'thinking' field, and final messages (the string between '<|message|>' and '<|end|>') in the 'content' field.") }}
+            {%- endif %}
+        {%- endif %}
+        {%- if "thinking" in message %}
+            {%- if "<|channel|>analysis<|message|>" in message.thinking or "<|channel|>final<|message|>" in message.thinking %}
+                {{- raise_exception("You have passed a message containing <|channel|> tags in the thinking field. Instead of doing this, you should pass analysis messages (the string between '<|message|>' and '<|end|>') in the 'thinking' field, and final messages (the string between '<|message|>' and '<|end|>') in the 'content' field.") }}
+            {%- endif %}
+        {%- endif %}
+        {%- if "tool_calls" in message %}
+            {#- We need very careful handling here - we want to drop the tool call analysis message if the model #}
+            {#- has output a later <|final|> message, but otherwise we want to retain it. This is the only case #}
+            {#- when we render CoT/analysis messages in inference. #}
+            {%- set future_final_message = namespace(found=false) %}
+            {%- for future_message in loop_messages[loop.index:] %}
+                {%- if future_message.role == 'assistant' and "tool_calls" not in future_message %}
+                    {%- set future_final_message.found = true %}
+                {%- endif %}
+            {%- endfor %}
+            {#- We assume max 1 tool call per message, and so we infer the tool call name #}
+            {#- in "tool" messages from the most recent assistant tool call name #}
+            {%- set tool_call = message.tool_calls[0] %}
+            {%- if tool_call.function %}
+                {%- set tool_call = tool_call.function %}
+            {%- endif %}
+            {%- if message.content and message.thinking %}
+                {{- raise_exception("Cannot pass both content and thinking in an assistant message with tool calls! Put the analysis message in one or the other, but not both.") }}
+            {%- elif message.content and not future_final_message.found %}
+                {{- "<|start|>assistant<|channel|>analysis<|message|>" + message.content + "<|end|>" }}
+            {%- elif message.thinking and not future_final_message.found %}
+                {{- "<|start|>assistant<|channel|>analysis<|message|>" + message.thinking + "<|end|>" }}
+            {%- endif %}
+            {{- "<|start|>assistant to=" }}
+            {{- "functions." + tool_call.name + "<|channel|>commentary " }}
+            {{- (tool_call.content_type if tool_call.content_type is defined else "json") + "<|message|>" }}
+            {{- tool_call.arguments|tojson }}
+            {{- "<|call|>" }}
+            {%- set last_tool_call.name = tool_call.name %}
+        {%- elif loop.last and not add_generation_prompt %}
+            {#- Only render the CoT if the final turn is an assistant turn and add_generation_prompt is false #}
+            {#- This is a situation that should only occur in training, never in inference. #}
+            {%- if "thinking" in message %}
+                {{- "<|start|>assistant<|channel|>analysis<|message|>" + message.thinking + "<|end|>" }}
+            {%- endif %}
+            {#- <|return|> indicates the end of generation, but <|end|> does not #}
+            {#- <|return|> should never be an input to the model, but we include it as the final token #}
+            {#- when training, so the model learns to emit it. #}
+            {{- "<|start|>assistant<|channel|>final<|message|>" + message.content + "<|return|>" }}
+        {%- else %}
+            {#- CoT is dropped during all previous turns, so we never render it for inference #}
+            {{- "<|start|>assistant<|channel|>final<|message|>" + message.content + "<|end|>" }}
+            {%- set last_tool_call.name = none %}
+        {%- endif %}
+    {%- elif message.role == 'tool' -%}
+        {%- if last_tool_call.name is none %}
+            {{- raise_exception("Message has tool role, but there was no previous assistant message with a tool call!") }}
+        {%- endif %}
+        {{- "<|start|>functions." + last_tool_call.name }}
+        {{- " to=assistant<|channel|>commentary<|message|>" + message.content|tojson + "<|end|>" }}
+    {%- elif message.role == 'user' -%}
+        {{- "<|start|>user<|message|>" + message.content + "<|end|>" }}
+    {%- endif -%}
+{%- endfor -%}
+
+{#- Generation prompt #}
+{%- if add_generation_prompt -%}
+<|start|>assistant
+{%- endif -%}

citation.json

@@ -0,0 +1,9 @@
+{
+  "title": "GPT-OSS MoE Expert Fingerprinting: Analyzing Expert Activation Patterns in Mixture of Experts Models",
+  "authors": [
+    "Aman Priyanshu",
+    "Supriti Vijay"
+  ],
+  "year": 2025,
+  "url": "https://amanpriyanshu.github.io/GPT-OSS-MoE-ExpertFingerprinting/"
+}

config.json

@@ -0,0 +1,123 @@
+{
+  "vocab_size": 201088,
+  "hidden_size": 2880,
+  "intermediate_size": 2880,
+  "num_hidden_layers": 24,
+  "num_attention_heads": 64,
+  "num_local_experts": 21,
+  "sliding_window": 128,
+  "num_experts_per_tok": 4,
+  "num_key_value_heads": 8,
+  "hidden_act": "silu",
+  "initializer_range": 0.02,
+  "rms_norm_eps": 1e-05,
+  "rope_theta": 150000,
+  "rope_scaling": {
+    "beta_fast": 32.0,
+    "beta_slow": 1.0,
+    "factor": 32.0,
+    "original_max_position_embeddings": 4096,
+    "rope_type": "yarn",
+    "truncate": false
+  },
+  "attention_dropout": 0.0,
+  "head_dim": 64,
+  "layer_types": [
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "full_attention"
+  ],
+  "attention_bias": true,
+  "max_position_embeddings": 131072,
+  "router_aux_loss_coef": 0.9,
+  "output_router_logits": false,
+  "use_cache": true,
+  "return_dict": true,
+  "output_hidden_states": false,
+  "torchscript": false,
+  "torch_dtype": null,
+  "pruned_heads": {},
+  "tie_word_embeddings": false,
+  "chunk_size_feed_forward": 0,
+  "is_encoder_decoder": false,
+  "is_decoder": false,
+  "cross_attention_hidden_size": null,
+  "add_cross_attention": false,
+  "tie_encoder_decoder": false,
+  "architectures": [
+    "GptOssForCausalLM"
+  ],
+  "finetuning_task": null,
+  "id2label": {
+    "0": "LABEL_0",
+    "1": "LABEL_1"
+  },
+  "label2id": {
+    "LABEL_0": 0,
+    "LABEL_1": 1
+  },
+  "task_specific_params": null,
+  "problem_type": null,
+  "tokenizer_class": null,
+  "prefix": null,
+  "bos_token_id": null,
+  "pad_token_id": 199999,
+  "eos_token_id": 200002,
+  "sep_token_id": null,
+  "decoder_start_token_id": null,
+  "max_length": 20,
+  "min_length": 0,
+  "do_sample": false,
+  "early_stopping": false,
+  "num_beams": 1,
+  "num_beam_groups": 1,
+  "diversity_penalty": 0.0,
+  "temperature": 1.0,
+  "top_k": 50,
+  "top_p": 1.0,
+  "typical_p": 1.0,
+  "repetition_penalty": 1.0,
+  "length_penalty": 1.0,
+  "no_repeat_ngram_size": 0,
+  "encoder_no_repeat_ngram_size": 0,
+  "bad_words_ids": null,
+  "num_return_sequences": 1,
+  "output_scores": false,
+  "return_dict_in_generate": false,
+  "forced_bos_token_id": null,
+  "forced_eos_token_id": null,
+  "remove_invalid_values": false,
+  "exponential_decay_length_penalty": null,
+  "suppress_tokens": null,
+  "begin_suppress_tokens": null,
+  "_name_or_path": "openai/gpt-oss-20b",
+  "transformers_version": "4.55.0",
+  "experts_per_token": 4,
+  "initial_context_length": 4096,
+  "model_type": "gpt_oss",
+  "swiglu_limit": 7.0,
+  "tf_legacy_loss": false,
+  "use_bfloat16": false,
+  "output_attentions": false
+}

expert_mapping.json

@@ -0,0 +1,554 @@
+{
+  "0": {
+    "11": 0,
+    "9": 1,
+    "16": 2,
+    "3": 3,
+    "1": 4,
+    "17": 5,
+    "24": 6,
+    "22": 7,
+    "2": 8,
+    "31": 9,
+    "7": 10,
+    "25": 11,
+    "28": 12,
+    "5": 13,
+    "6": 14,
+    "13": 15,
+    "23": 16,
+    "10": 17,
+    "21": 18,
+    "30": 19,
+    "8": 20
+  },
+  "1": {
+    "0": 0,
+    "28": 1,
+    "25": 2,
+    "27": 3,
+    "1": 4,
+    "10": 5,
+    "18": 6,
+    "16": 7,
+    "6": 8,
+    "12": 9,
+    "26": 10,
+    "13": 11,
+    "29": 12,
+    "20": 13,
+    "17": 14,
+    "30": 15,
+    "11": 16,
+    "23": 17,
+    "24": 18,
+    "7": 19,
+    "22": 20
+  },
+  "2": {
+    "21": 0,
+    "31": 1,
+    "5": 2,
+    "29": 3,
+    "0": 4,
+    "10": 5,
+    "8": 6,
+    "22": 7,
+    "3": 8,
+    "28": 9,
+    "6": 10,
+    "20": 11,
+    "4": 12,
+    "27": 13,
+    "16": 14,
+    "24": 15,
+    "25": 16,
+    "15": 17,
+    "7": 18,
+    "14": 19,
+    "26": 20
+  },
+  "3": {
+    "31": 0,
+    "7": 1,
+    "11": 2,
+    "0": 3,
+    "2": 4,
+    "23": 5,
+    "19": 6,
+    "21": 7,
+    "12": 8,
+    "25": 9,
+    "22": 10,
+    "1": 11,
+    "3": 12,
+    "27": 13,
+    "16": 14,
+    "9": 15,
+    "24": 16,
+    "18": 17,
+    "5": 18,
+    "13": 19,
+    "6": 20
+  },
+  "4": {
+    "6": 0,
+    "18": 1,
+    "15": 2,
+    "14": 3,
+    "20": 4,
+    "30": 5,
+    "17": 6,
+    "24": 7,
+    "12": 8,
+    "23": 9,
+    "10": 10,
+    "4": 11,
+    "7": 12,
+    "31": 13,
+    "21": 14,
+    "27": 15,
+    "13": 16,
+    "25": 17,
+    "8": 18,
+    "1": 19,
+    "22": 20
+  },
+  "5": {
+    "6": 0,
+    "31": 1,
+    "0": 2,
+    "10": 3,
+    "17": 4,
+    "14": 5,
+    "24": 6,
+    "15": 7,
+    "2": 8,
+    "18": 9,
+    "5": 10,
+    "27": 11,
+    "28": 12,
+    "8": 13,
+    "16": 14,
+    "30": 15,
+    "21": 16,
+    "12": 17,
+    "26": 18,
+    "4": 19,
+    "11": 20
+  },
+  "6": {
+    "16": 0,
+    "2": 1,
+    "10": 2,
+    "9": 3,
+    "31": 4,
+    "3": 5,
+    "5": 6,
+    "20": 7,
+    "4": 8,
+    "7": 9,
+    "15": 10,
+    "27": 11,
+    "28": 12,
+    "11": 13,
+    "14": 14,
+    "21": 15,
+    "18": 16,
+    "24": 17,
+    "30": 18,
+    "8": 19,
+    "26": 20
+  },
+  "7": {
+    "15": 0,
+    "5": 1,
+    "14": 2,
+    "1": 3,
+    "30": 4,
+    "24": 5,
+    "13": 6,
+    "10": 7,
+    "29": 8,
+    "28": 9,
+    "4": 10,
+    "17": 11,
+    "21": 12,
+    "19": 13,
+    "3": 14,
+    "20": 15,
+    "16": 16,
+    "9": 17,
+    "11": 18,
+    "23": 19,
+    "0": 20
+  },
+  "8": {
+    "23": 0,
+    "24": 1,
+    "9": 2,
+    "20": 3,
+    "4": 4,
+    "0": 5,
+    "17": 6,
+    "31": 7,
+    "30": 8,
+    "5": 9,
+    "3": 10,
+    "19": 11,
+    "13": 12,
+    "28": 13,
+    "10": 14,
+    "6": 15,
+    "26": 16,
+    "8": 17,
+    "7": 18,
+    "2": 19,
+    "1": 20
+  },
+  "9": {
+    "8": 0,
+    "26": 1,
+    "19": 2,
+    "13": 3,
+    "15": 4,
+    "30": 5,
+    "16": 6,
+    "14": 7,
+    "22": 8,
+    "5": 9,
+    "21": 10,
+    "23": 11,
+    "4": 12,
+    "3": 13,
+    "29": 14,
+    "20": 15,
+    "2": 16,
+    "1": 17,
+    "31": 18,
+    "17": 19,
+    "0": 20
+  },
+  "10": {
+    "10": 0,
+    "5": 1,
+    "23": 2,
+    "7": 3,
+    "21": 4,
+    "26": 5,
+    "19": 6,
+    "18": 7,
+    "8": 8,
+    "11": 9,
+    "0": 10,
+    "24": 11,
+    "25": 12,
+    "30": 13,
+    "2": 14,
+    "15": 15,
+    "16": 16,
+    "27": 17,
+    "14": 18,
+    "28": 19,
+    "13": 20
+  },
+  "11": {
+    "30": 0,
+    "18": 1,
+    "9": 2,
+    "11": 3,
+    "27": 4,
+    "20": 5,
+    "21": 6,
+    "7": 7,
+    "25": 8,
+    "17": 9,
+    "4": 10,
+    "24": 11,
+    "13": 12,
+    "15": 13,
+    "23": 14,
+    "12": 15,
+    "5": 16,
+    "26": 17,
+    "29": 18,
+    "22": 19,
+    "1": 20
+  },
+  "12": {
+    "23": 0,
+    "3": 1,
+    "14": 2,
+    "28": 3,
+    "21": 4,
+    "12": 5,
+    "25": 6,
+    "1": 7,
+    "24": 8,
+    "18": 9,
+    "17": 10,
+    "6": 11,
+    "8": 12,
+    "7": 13,
+    "30": 14,
+    "13": 15,
+    "19": 16,
+    "26": 17,
+    "0": 18,
+    "2": 19,
+    "11": 20
+  },
+  "13": {
+    "15": 0,
+    "21": 1,
+    "16": 2,
+    "9": 3,
+    "11": 4,
+    "8": 5,
+    "1": 6,
+    "5": 7,
+    "19": 8,
+    "24": 9,
+    "3": 10,
+    "30": 11,
+    "7": 12,
+    "20": 13,
+    "10": 14,
+    "0": 15,
+    "4": 16,
+    "29": 17,
+    "28": 18,
+    "12": 19,
+    "27": 20
+  },
+  "14": {
+    "6": 0,
+    "9": 1,
+    "31": 2,
+    "0": 3,
+    "21": 4,
+    "18": 5,
+    "23": 6,
+    "24": 7,
+    "26": 8,
+    "4": 9,
+    "11": 10,
+    "7": 11,
+    "1": 12,
+    "27": 13,
+    "3": 14,
+    "29": 15,
+    "17": 16,
+    "12": 17,
+    "16": 18,
+    "13": 19,
+    "22": 20
+  },
+  "15": {
+    "22": 0,
+    "13": 1,
+    "19": 2,
+    "18": 3,
+    "3": 4,
+    "31": 5,
+    "26": 6,
+    "2": 7,
+    "9": 8,
+    "28": 9,
+    "29": 10,
+    "6": 11,
+    "10": 12,
+    "0": 13,
+    "4": 14,
+    "25": 15,
+    "14": 16,
+    "8": 17,
+    "7": 18,
+    "27": 19,
+    "23": 20
+  },
+  "16": {
+    "31": 0,
+    "10": 1,
+    "30": 2,
+    "5": 3,
+    "15": 4,
+    "11": 5,
+    "24": 6,
+    "17": 7,
+    "0": 8,
+    "20": 9,
+    "18": 10,
+    "21": 11,
+    "2": 12,
+    "1": 13,
+    "4": 14,
+    "12": 15,
+    "9": 16,
+    "3": 17,
+    "14": 18,
+    "22": 19,
+    "13": 20
+  },
+  "17": {
+    "28": 0,
+    "14": 1,
+    "11": 2,
+    "30": 3,
+    "13": 4,
+    "5": 5,
+    "17": 6,
+    "4": 7,
+    "7": 8,
+    "19": 9,
+    "8": 10,
+    "25": 11,
+    "12": 12,
+    "1": 13,
+    "2": 14,
+    "20": 15,
+    "24": 16,
+    "16": 17,
+    "0": 18,
+    "15": 19,
+    "22": 20
+  },
+  "18": {
+    "1": 0,
+    "9": 1,
+    "18": 2,
+    "14": 3,
+    "24": 4,
+    "23": 5,
+    "27": 6,
+    "15": 7,
+    "20": 8,
+    "2": 9,
+    "6": 10,
+    "3": 11,
+    "26": 12,
+    "7": 13,
+    "31": 14,
+    "13": 15,
+    "19": 16,
+    "16": 17,
+    "5": 18,
+    "12": 19,
+    "11": 20
+  },
+  "19": {
+    "16": 0,
+    "3": 1,
+    "6": 2,
+    "19": 3,
+    "29": 4,
+    "12": 5,
+    "20": 6,
+    "31": 7,
+    "2": 8,
+    "7": 9,
+    "9": 10,
+    "0": 11,
+    "17": 12,
+    "13": 13,
+    "25": 14,
+    "4": 15,
+    "15": 16,
+    "22": 17,
+    "10": 18,
+    "11": 19,
+    "1": 20
+  },
+  "20": {
+    "21": 0,
+    "22": 1,
+    "0": 2,
+    "16": 3,
+    "30": 4,
+    "13": 5,
+    "24": 6,
+    "25": 7,
+    "31": 8,
+    "11": 9,
+    "6": 10,
+    "12": 11,
+    "20": 12,
+    "10": 13,
+    "7": 14,
+    "1": 15,
+    "23": 16,
+    "15": 17,
+    "5": 18,
+    "4": 19,
+    "26": 20
+  },
+  "21": {
+    "3": 0,
+    "9": 1,
+    "31": 2,
+    "27": 3,
+    "1": 4,
+    "21": 5,
+    "2": 6,
+    "6": 7,
+    "18": 8,
+    "20": 9,
+    "22": 10,
+    "13": 11,
+    "10": 12,
+    "4": 13,
+    "14": 14,
+    "25": 15,
+    "30": 16,
+    "11": 17,
+    "29": 18,
+    "7": 19,
+    "23": 20
+  },
+  "22": {
+    "25": 0,
+    "8": 1,
+    "16": 2,
+    "18": 3,
+    "14": 4,
+    "4": 5,
+    "0": 6,
+    "22": 7,
+    "15": 8,
+    "13": 9,
+    "7": 10,
+    "20": 11,
+    "27": 12,
+    "23": 13,
+    "26": 14,
+    "10": 15,
+    "24": 16,
+    "19": 17,
+    "31": 18,
+    "5": 19,
+    "11": 20
+  },
+  "23": {
+    "31": 0,
+    "18": 1,
+    "0": 2,
+    "10": 3,
+    "16": 4,
+    "22": 5,
+    "27": 6,
+    "15": 7,
+    "28": 8,
+    "5": 9,
+    "17": 10,
+    "7": 11,
+    "8": 12,
+    "3": 13,
+    "23": 14,
+    "14": 15,
+    "19": 16,
+    "25": 17,
+    "13": 18,
+    "24": 19,
+    "4": 20
+  }
+}

generation_config.json

@@ -0,0 +1,10 @@
+{
+  "bos_token_id": 199998,
+  "do_sample": true,
+  "eos_token_id": [
+    200002,
+    199999
+  ],
+  "pad_token_id": 199999,
+  "transformers_version": "4.55.0"
+}

model-00001-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1bee9b34efb6d934f0f8ab926f790cb8184a210b514d57fc88f5abd7ea041717
+size 1908461290

model-00002-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d57636338fd18fde34a8780cb1a4e5d838d239206b0cb0daf533d4d19bc9462a
+size 1500412284

model-00003-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:725ee6a55cc0718c654de38377c80345b712b57da40d831b3e4d1c379cf4eea7
+size 1795663794

model-00004-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba736bab28c4fca830ee42922c17d92f671fa0027f7974fd96cb5b20cf3f3b2b
+size 1500412284

model-00005-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d58872958c811224081558cc5d2673e9f11ca8647a6ff35a1f9d66a985f7c928
+size 1795663794

model-00006-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bbf91e0d67904770b0567e22a2e85ff48936a1837be593dc0ae65b6a76eac2f6
+size 1500412284

model-00007-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e67a410420574f31db890b57619d896f055f0b87b6928fdbada8b0ff89847f24
+size 1795663794

model-00008-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b74aac309aba676d5205bd1642905cc98ffc1a9250223d0087371cc1da5b8e33
+size 1500412316

model-00009-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da1ddf9125be1950ac194193cfc58652555c91c34f06a28ff05455d7a99eb208
+size 1795663818

model-00010-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a154279c5120aaf6f73ea7a655d2152065b6ac0fb0416a14a09057af8cc433e
+size 1500412316

model-00011-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a21423f21af0fe8d088300a3b0f429465f62d6dc6f21dca66887caef1e7e0b30
+size 1795663818

model-00012-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd37133f9ca132e87c1fcdfe0d7880729d64d3822dc50af4f86e5f7879b3017c
+size 1500412316

model-00013-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd2125da3c6c2aa82b9662fc7915a6fe8f578edc0cfa8b16be5658a610cdf857
+size 1795663818

model-00014-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0ee5bf57f3c18bd54014d8ae257b5166ac108c06dd930b74a62062e3cd6ee0c
+size 1500412316

model-00015-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd518c0f16f048a6fda64fb306d71b00e3175c42229bd30b6eb069e9a1359467
+size 1795663818

model-00016-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9c2dd462d29b786a0a595288195e311f70fa50028f100c0f8ac14065920b74b
+size 1500412316

model-00017-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a22598447572c82c26a33c83934356d3c5fbfde6de2cc99923dc507257c82bcc
+size 1045475384

model-00018-of-00018.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:08e257de1fa16cf80c8a36ca29c5974fc999f0dc7ada1468bfd0be18151af482
+size 1158267008

model.safetensors.index.json

@@ -0,0 +1,419 @@
+{
+  "metadata": {
+    "total_parameters": 14342550840,
+    "total_size": 28685101680
+  },
+  "weight_map": {
+    "lm_head.weight": "model-00018-of-00018.safetensors",
+    "model.embed_tokens.weight": "model-00001-of-00018.safetensors",
+    "model.layers.0.input_layernorm.weight": "model-00002-of-00018.safetensors",
+    "model.layers.0.mlp.experts.down_proj": "model-00002-of-00018.safetensors",
+    "model.layers.0.mlp.experts.down_proj_bias": "model-00002-of-00018.safetensors",
+    "model.layers.0.mlp.experts.gate_up_proj": "model-00001-of-00018.safetensors",
+    "model.layers.0.mlp.experts.gate_up_proj_bias": "model-00001-of-00018.safetensors",
+    "model.layers.0.mlp.router.bias": "model-00001-of-00018.safetensors",
+    "model.layers.0.mlp.router.weight": "model-00001-of-00018.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "model-00002-of-00018.safetensors",
+    "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00018.safetensors",
+    "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00018.safetensors",
+    "model.layers.0.self_attn.o_proj.bias": "model-00001-of-00018.safetensors",
+    "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00018.safetensors",
+    "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00018.safetensors",
+    "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00018.safetensors",
+    "model.layers.0.self_attn.sinks": "model-00001-of-00018.safetensors",
+    "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00018.safetensors",
+    "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00018.safetensors",
+    "model.layers.1.input_layernorm.weight": "model-00002-of-00018.safetensors",
+    "model.layers.1.mlp.experts.down_proj": "model-00002-of-00018.safetensors",
+    "model.layers.1.mlp.experts.down_proj_bias": "model-00002-of-00018.safetensors",
+    "model.layers.1.mlp.experts.gate_up_proj": "model-00002-of-00018.safetensors",
+    "model.layers.1.mlp.experts.gate_up_proj_bias": "model-00002-of-00018.safetensors",
+    "model.layers.1.mlp.router.bias": "model-00002-of-00018.safetensors",
+    "model.layers.1.mlp.router.weight": "model-00002-of-00018.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "model-00002-of-00018.safetensors",
+    "model.layers.1.self_attn.k_proj.bias": "model-00002-of-00018.safetensors",
+    "model.layers.1.self_attn.k_proj.weight": "model-00002-of-00018.safetensors",
+    "model.layers.1.self_attn.o_proj.bias": "model-00002-of-00018.safetensors",
+    "model.layers.1.self_attn.o_proj.weight": "model-00002-of-00018.safetensors",
+    "model.layers.1.self_attn.q_proj.bias": "model-00002-of-00018.safetensors",
+    "model.layers.1.self_attn.q_proj.weight": "model-00002-of-00018.safetensors",
+    "model.layers.1.self_attn.sinks": "model-00002-of-00018.safetensors",
+    "model.layers.1.self_attn.v_proj.bias": "model-00002-of-00018.safetensors",
+    "model.layers.1.self_attn.v_proj.weight": "model-00002-of-00018.safetensors",
+    "model.layers.10.input_layernorm.weight": "model-00008-of-00018.safetensors",
+    "model.layers.10.mlp.experts.down_proj": "model-00008-of-00018.safetensors",
+    "model.layers.10.mlp.experts.down_proj_bias": "model-00008-of-00018.safetensors",
+    "model.layers.10.mlp.experts.gate_up_proj": "model-00008-of-00018.safetensors",
+    "model.layers.10.mlp.experts.gate_up_proj_bias": "model-00008-of-00018.safetensors",
+    "model.layers.10.mlp.router.bias": "model-00008-of-00018.safetensors",
+    "model.layers.10.mlp.router.weight": "model-00008-of-00018.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "model-00008-of-00018.safetensors",
+    "model.layers.10.self_attn.k_proj.bias": "model-00008-of-00018.safetensors",
+    "model.layers.10.self_attn.k_proj.weight": "model-00008-of-00018.safetensors",
+    "model.layers.10.self_attn.o_proj.bias": "model-00008-of-00018.safetensors",
+    "model.layers.10.self_attn.o_proj.weight": "model-00008-of-00018.safetensors",
+    "model.layers.10.self_attn.q_proj.bias": "model-00008-of-00018.safetensors",
+    "model.layers.10.self_attn.q_proj.weight": "model-00008-of-00018.safetensors",
+    "model.layers.10.self_attn.sinks": "model-00008-of-00018.safetensors",
+    "model.layers.10.self_attn.v_proj.bias": "model-00008-of-00018.safetensors",
+    "model.layers.10.self_attn.v_proj.weight": "model-00008-of-00018.safetensors",
+    "model.layers.11.input_layernorm.weight": "model-00009-of-00018.safetensors",
+    "model.layers.11.mlp.experts.down_proj": "model-00009-of-00018.safetensors",
+    "model.layers.11.mlp.experts.down_proj_bias": "model-00009-of-00018.safetensors",
+    "model.layers.11.mlp.experts.gate_up_proj": "model-00009-of-00018.safetensors",
+    "model.layers.11.mlp.experts.gate_up_proj_bias": "model-00009-of-00018.safetensors",
+    "model.layers.11.mlp.router.bias": "model-00008-of-00018.safetensors",
+    "model.layers.11.mlp.router.weight": "model-00008-of-00018.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "model-00009-of-00018.safetensors",
+    "model.layers.11.self_attn.k_proj.bias": "model-00008-of-00018.safetensors",
+    "model.layers.11.self_attn.k_proj.weight": "model-00008-of-00018.safetensors",
+    "model.layers.11.self_attn.o_proj.bias": "model-00008-of-00018.safetensors",
+    "model.layers.11.self_attn.o_proj.weight": "model-00008-of-00018.safetensors",
+    "model.layers.11.self_attn.q_proj.bias": "model-00008-of-00018.safetensors",
+    "model.layers.11.self_attn.q_proj.weight": "model-00008-of-00018.safetensors",
+    "model.layers.11.self_attn.sinks": "model-00008-of-00018.safetensors",
+    "model.layers.11.self_attn.v_proj.bias": "model-00008-of-00018.safetensors",
+    "model.layers.11.self_attn.v_proj.weight": "model-00008-of-00018.safetensors",
+    "model.layers.12.input_layernorm.weight": "model-00010-of-00018.safetensors",
+    "model.layers.12.mlp.experts.down_proj": "model-00010-of-00018.safetensors",
+    "model.layers.12.mlp.experts.down_proj_bias": "model-00010-of-00018.safetensors",
+    "model.layers.12.mlp.experts.gate_up_proj": "model-00009-of-00018.safetensors",
+    "model.layers.12.mlp.experts.gate_up_proj_bias": "model-00009-of-00018.safetensors",
+    "model.layers.12.mlp.router.bias": "model-00009-of-00018.safetensors",
+    "model.layers.12.mlp.router.weight": "model-00009-of-00018.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "model-00010-of-00018.safetensors",
+    "model.layers.12.self_attn.k_proj.bias": "model-00009-of-00018.safetensors",
+    "model.layers.12.self_attn.k_proj.weight": "model-00009-of-00018.safetensors",
+    "model.layers.12.self_attn.o_proj.bias": "model-00009-of-00018.safetensors",
+    "model.layers.12.self_attn.o_proj.weight": "model-00009-of-00018.safetensors",
+    "model.layers.12.self_attn.q_proj.bias": "model-00009-of-00018.safetensors",
+    "model.layers.12.self_attn.q_proj.weight": "model-00009-of-00018.safetensors",
+    "model.layers.12.self_attn.sinks": "model-00009-of-00018.safetensors",
+    "model.layers.12.self_attn.v_proj.bias": "model-00009-of-00018.safetensors",
+    "model.layers.12.self_attn.v_proj.weight": "model-00009-of-00018.safetensors",
+    "model.layers.13.input_layernorm.weight": "model-00010-of-00018.safetensors",
+    "model.layers.13.mlp.experts.down_proj": "model-00010-of-00018.safetensors",
+    "model.layers.13.mlp.experts.down_proj_bias": "model-00010-of-00018.safetensors",
+    "model.layers.13.mlp.experts.gate_up_proj": "model-00010-of-00018.safetensors",
+    "model.layers.13.mlp.experts.gate_up_proj_bias": "model-00010-of-00018.safetensors",
+    "model.layers.13.mlp.router.bias": "model-00010-of-00018.safetensors",
+    "model.layers.13.mlp.router.weight": "model-00010-of-00018.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "model-00010-of-00018.safetensors",
+    "model.layers.13.self_attn.k_proj.bias": "model-00010-of-00018.safetensors",
+    "model.layers.13.self_attn.k_proj.weight": "model-00010-of-00018.safetensors",
+    "model.layers.13.self_attn.o_proj.bias": "model-00010-of-00018.safetensors",
+    "model.layers.13.self_attn.o_proj.weight": "model-00010-of-00018.safetensors",
+    "model.layers.13.self_attn.q_proj.bias": "model-00010-of-00018.safetensors",
+    "model.layers.13.self_attn.q_proj.weight": "model-00010-of-00018.safetensors",
+    "model.layers.13.self_attn.sinks": "model-00010-of-00018.safetensors",
+    "model.layers.13.self_attn.v_proj.bias": "model-00010-of-00018.safetensors",
+    "model.layers.13.self_attn.v_proj.weight": "model-00010-of-00018.safetensors",
+    "model.layers.14.input_layernorm.weight": "model-00011-of-00018.safetensors",
+    "model.layers.14.mlp.experts.down_proj": "model-00011-of-00018.safetensors",
+    "model.layers.14.mlp.experts.down_proj_bias": "model-00011-of-00018.safetensors",
+    "model.layers.14.mlp.experts.gate_up_proj": "model-00011-of-00018.safetensors",
+    "model.layers.14.mlp.experts.gate_up_proj_bias": "model-00011-of-00018.safetensors",
+    "model.layers.14.mlp.router.bias": "model-00010-of-00018.safetensors",
+    "model.layers.14.mlp.router.weight": "model-00010-of-00018.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "model-00011-of-00018.safetensors",
+    "model.layers.14.self_attn.k_proj.bias": "model-00010-of-00018.safetensors",
+    "model.layers.14.self_attn.k_proj.weight": "model-00010-of-00018.safetensors",
+    "model.layers.14.self_attn.o_proj.bias": "model-00010-of-00018.safetensors",
+    "model.layers.14.self_attn.o_proj.weight": "model-00010-of-00018.safetensors",
+    "model.layers.14.self_attn.q_proj.bias": "model-00010-of-00018.safetensors",
+    "model.layers.14.self_attn.q_proj.weight": "model-00010-of-00018.safetensors",
+    "model.layers.14.self_attn.sinks": "model-00010-of-00018.safetensors",
+    "model.layers.14.self_attn.v_proj.bias": "model-00010-of-00018.safetensors",
+    "model.layers.14.self_attn.v_proj.weight": "model-00010-of-00018.safetensors",
+    "model.layers.15.input_layernorm.weight": "model-00012-of-00018.safetensors",
+    "model.layers.15.mlp.experts.down_proj": "model-00012-of-00018.safetensors",
+    "model.layers.15.mlp.experts.down_proj_bias": "model-00012-of-00018.safetensors",
+    "model.layers.15.mlp.experts.gate_up_proj": "model-00011-of-00018.safetensors",
+    "model.layers.15.mlp.experts.gate_up_proj_bias": "model-00011-of-00018.safetensors",
+    "model.layers.15.mlp.router.bias": "model-00011-of-00018.safetensors",
+    "model.layers.15.mlp.router.weight": "model-00011-of-00018.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "model-00012-of-00018.safetensors",
+    "model.layers.15.self_attn.k_proj.bias": "model-00011-of-00018.safetensors",
+    "model.layers.15.self_attn.k_proj.weight": "model-00011-of-00018.safetensors",
+    "model.layers.15.self_attn.o_proj.bias": "model-00011-of-00018.safetensors",
+    "model.layers.15.self_attn.o_proj.weight": "model-00011-of-00018.safetensors",
+    "model.layers.15.self_attn.q_proj.bias": "model-00011-of-00018.safetensors",
+    "model.layers.15.self_attn.q_proj.weight": "model-00011-of-00018.safetensors",
+    "model.layers.15.self_attn.sinks": "model-00011-of-00018.safetensors",
+    "model.layers.15.self_attn.v_proj.bias": "model-00011-of-00018.safetensors",
+    "model.layers.15.self_attn.v_proj.weight": "model-00011-of-00018.safetensors",
+    "model.layers.16.input_layernorm.weight": "model-00012-of-00018.safetensors",
+    "model.layers.16.mlp.experts.down_proj": "model-00012-of-00018.safetensors",
+    "model.layers.16.mlp.experts.down_proj_bias": "model-00012-of-00018.safetensors",
+    "model.layers.16.mlp.experts.gate_up_proj": "model-00012-of-00018.safetensors",
+    "model.layers.16.mlp.experts.gate_up_proj_bias": "model-00012-of-00018.safetensors",
+    "model.layers.16.mlp.router.bias": "model-00012-of-00018.safetensors",
+    "model.layers.16.mlp.router.weight": "model-00012-of-00018.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "model-00012-of-00018.safetensors",
+    "model.layers.16.self_attn.k_proj.bias": "model-00012-of-00018.safetensors",
+    "model.layers.16.self_attn.k_proj.weight": "model-00012-of-00018.safetensors",
+    "model.layers.16.self_attn.o_proj.bias": "model-00012-of-00018.safetensors",
+    "model.layers.16.self_attn.o_proj.weight": "model-00012-of-00018.safetensors",
+    "model.layers.16.self_attn.q_proj.bias": "model-00012-of-00018.safetensors",
+    "model.layers.16.self_attn.q_proj.weight": "model-00012-of-00018.safetensors",
+    "model.layers.16.self_attn.sinks": "model-00012-of-00018.safetensors",
+    "model.layers.16.self_attn.v_proj.bias": "model-00012-of-00018.safetensors",
+    "model.layers.16.self_attn.v_proj.weight": "model-00012-of-00018.safetensors",
+    "model.layers.17.input_layernorm.weight": "model-00013-of-00018.safetensors",
+    "model.layers.17.mlp.experts.down_proj": "model-00013-of-00018.safetensors",
+    "model.layers.17.mlp.experts.down_proj_bias": "model-00013-of-00018.safetensors",
+    "model.layers.17.mlp.experts.gate_up_proj": "model-00013-of-00018.safetensors",
+    "model.layers.17.mlp.experts.gate_up_proj_bias": "model-00013-of-00018.safetensors",
+    "model.layers.17.mlp.router.bias": "model-00012-of-00018.safetensors",
+    "model.layers.17.mlp.router.weight": "model-00012-of-00018.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "model-00013-of-00018.safetensors",
+    "model.layers.17.self_attn.k_proj.bias": "model-00012-of-00018.safetensors",
+    "model.layers.17.self_attn.k_proj.weight": "model-00012-of-00018.safetensors",
+    "model.layers.17.self_attn.o_proj.bias": "model-00012-of-00018.safetensors",
+    "model.layers.17.self_attn.o_proj.weight": "model-00012-of-00018.safetensors",
+    "model.layers.17.self_attn.q_proj.bias": "model-00012-of-00018.safetensors",
+    "model.layers.17.self_attn.q_proj.weight": "model-00012-of-00018.safetensors",
+    "model.layers.17.self_attn.sinks": "model-00012-of-00018.safetensors",
+    "model.layers.17.self_attn.v_proj.bias": "model-00012-of-00018.safetensors",
+    "model.layers.17.self_attn.v_proj.weight": "model-00012-of-00018.safetensors",
+    "model.layers.18.input_layernorm.weight": "model-00014-of-00018.safetensors",
+    "model.layers.18.mlp.experts.down_proj": "model-00014-of-00018.safetensors",
+    "model.layers.18.mlp.experts.down_proj_bias": "model-00014-of-00018.safetensors",
+    "model.layers.18.mlp.experts.gate_up_proj": "model-00013-of-00018.safetensors",
+    "model.layers.18.mlp.experts.gate_up_proj_bias": "model-00013-of-00018.safetensors",
+    "model.layers.18.mlp.router.bias": "model-00013-of-00018.safetensors",
+    "model.layers.18.mlp.router.weight": "model-00013-of-00018.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "model-00014-of-00018.safetensors",
+    "model.layers.18.self_attn.k_proj.bias": "model-00013-of-00018.safetensors",
+    "model.layers.18.self_attn.k_proj.weight": "model-00013-of-00018.safetensors",
+    "model.layers.18.self_attn.o_proj.bias": "model-00013-of-00018.safetensors",
+    "model.layers.18.self_attn.o_proj.weight": "model-00013-of-00018.safetensors",
+    "model.layers.18.self_attn.q_proj.bias": "model-00013-of-00018.safetensors",
+    "model.layers.18.self_attn.q_proj.weight": "model-00013-of-00018.safetensors",
+    "model.layers.18.self_attn.sinks": "model-00013-of-00018.safetensors",
+    "model.layers.18.self_attn.v_proj.bias": "model-00013-of-00018.safetensors",
+    "model.layers.18.self_attn.v_proj.weight": "model-00013-of-00018.safetensors",
+    "model.layers.19.input_layernorm.weight": "model-00014-of-00018.safetensors",
+    "model.layers.19.mlp.experts.down_proj": "model-00014-of-00018.safetensors",
+    "model.layers.19.mlp.experts.down_proj_bias": "model-00014-of-00018.safetensors",
+    "model.layers.19.mlp.experts.gate_up_proj": "model-00014-of-00018.safetensors",
+    "model.layers.19.mlp.experts.gate_up_proj_bias": "model-00014-of-00018.safetensors",
+    "model.layers.19.mlp.router.bias": "model-00014-of-00018.safetensors",
+    "model.layers.19.mlp.router.weight": "model-00014-of-00018.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "model-00014-of-00018.safetensors",
+    "model.layers.19.self_attn.k_proj.bias": "model-00014-of-00018.safetensors",
+    "model.layers.19.self_attn.k_proj.weight": "model-00014-of-00018.safetensors",
+    "model.layers.19.self_attn.o_proj.bias": "model-00014-of-00018.safetensors",
+    "model.layers.19.self_attn.o_proj.weight": "model-00014-of-00018.safetensors",
+    "model.layers.19.self_attn.q_proj.bias": "model-00014-of-00018.safetensors",
+    "model.layers.19.self_attn.q_proj.weight": "model-00014-of-00018.safetensors",
+    "model.layers.19.self_attn.sinks": "model-00014-of-00018.safetensors",
+    "model.layers.19.self_attn.v_proj.bias": "model-00014-of-00018.safetensors",
+    "model.layers.19.self_attn.v_proj.weight": "model-00014-of-00018.safetensors",
+    "model.layers.2.input_layernorm.weight": "model-00003-of-00018.safetensors",
+    "model.layers.2.mlp.experts.down_proj": "model-00003-of-00018.safetensors",
+    "model.layers.2.mlp.experts.down_proj_bias": "model-00003-of-00018.safetensors",
+    "model.layers.2.mlp.experts.gate_up_proj": "model-00003-of-00018.safetensors",
+    "model.layers.2.mlp.experts.gate_up_proj_bias": "model-00003-of-00018.safetensors",
+    "model.layers.2.mlp.router.bias": "model-00002-of-00018.safetensors",
+    "model.layers.2.mlp.router.weight": "model-00002-of-00018.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "model-00003-of-00018.safetensors",
+    "model.layers.2.self_attn.k_proj.bias": "model-00002-of-00018.safetensors",
+    "model.layers.2.self_attn.k_proj.weight": "model-00002-of-00018.safetensors",
+    "model.layers.2.self_attn.o_proj.bias": "model-00002-of-00018.safetensors",
+    "model.layers.2.self_attn.o_proj.weight": "model-00002-of-00018.safetensors",
+    "model.layers.2.self_attn.q_proj.bias": "model-00002-of-00018.safetensors",
+    "model.layers.2.self_attn.q_proj.weight": "model-00002-of-00018.safetensors",
+    "model.layers.2.self_attn.sinks": "model-00002-of-00018.safetensors",
+    "model.layers.2.self_attn.v_proj.bias": "model-00002-of-00018.safetensors",
+    "model.layers.2.self_attn.v_proj.weight": "model-00002-of-00018.safetensors",
+    "model.layers.20.input_layernorm.weight": "model-00015-of-00018.safetensors",
+    "model.layers.20.mlp.experts.down_proj": "model-00015-of-00018.safetensors",
+    "model.layers.20.mlp.experts.down_proj_bias": "model-00015-of-00018.safetensors",
+    "model.layers.20.mlp.experts.gate_up_proj": "model-00015-of-00018.safetensors",
+    "model.layers.20.mlp.experts.gate_up_proj_bias": "model-00015-of-00018.safetensors",
+    "model.layers.20.mlp.router.bias": "model-00014-of-00018.safetensors",
+    "model.layers.20.mlp.router.weight": "model-00014-of-00018.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "model-00015-of-00018.safetensors",
+    "model.layers.20.self_attn.k_proj.bias": "model-00014-of-00018.safetensors",
+    "model.layers.20.self_attn.k_proj.weight": "model-00014-of-00018.safetensors",
+    "model.layers.20.self_attn.o_proj.bias": "model-00014-of-00018.safetensors",
+    "model.layers.20.self_attn.o_proj.weight": "model-00014-of-00018.safetensors",
+    "model.layers.20.self_attn.q_proj.bias": "model-00014-of-00018.safetensors",
+    "model.layers.20.self_attn.q_proj.weight": "model-00014-of-00018.safetensors",
+    "model.layers.20.self_attn.sinks": "model-00014-of-00018.safetensors",
+    "model.layers.20.self_attn.v_proj.bias": "model-00014-of-00018.safetensors",
+    "model.layers.20.self_attn.v_proj.weight": "model-00014-of-00018.safetensors",
+    "model.layers.21.input_layernorm.weight": "model-00016-of-00018.safetensors",
+    "model.layers.21.mlp.experts.down_proj": "model-00016-of-00018.safetensors",
+    "model.layers.21.mlp.experts.down_proj_bias": "model-00016-of-00018.safetensors",
+    "model.layers.21.mlp.experts.gate_up_proj": "model-00015-of-00018.safetensors",
+    "model.layers.21.mlp.experts.gate_up_proj_bias": "model-00015-of-00018.safetensors",
+    "model.layers.21.mlp.router.bias": "model-00015-of-00018.safetensors",
+    "model.layers.21.mlp.router.weight": "model-00015-of-00018.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "model-00016-of-00018.safetensors",
+    "model.layers.21.self_attn.k_proj.bias": "model-00015-of-00018.safetensors",
+    "model.layers.21.self_attn.k_proj.weight": "model-00015-of-00018.safetensors",
+    "model.layers.21.self_attn.o_proj.bias": "model-00015-of-00018.safetensors",
+    "model.layers.21.self_attn.o_proj.weight": "model-00015-of-00018.safetensors",
+    "model.layers.21.self_attn.q_proj.bias": "model-00015-of-00018.safetensors",
+    "model.layers.21.self_attn.q_proj.weight": "model-00015-of-00018.safetensors",
+    "model.layers.21.self_attn.sinks": "model-00015-of-00018.safetensors",
+    "model.layers.21.self_attn.v_proj.bias": "model-00015-of-00018.safetensors",
+    "model.layers.21.self_attn.v_proj.weight": "model-00015-of-00018.safetensors",
+    "model.layers.22.input_layernorm.weight": "model-00016-of-00018.safetensors",
+    "model.layers.22.mlp.experts.down_proj": "model-00016-of-00018.safetensors",
+    "model.layers.22.mlp.experts.down_proj_bias": "model-00016-of-00018.safetensors",
+    "model.layers.22.mlp.experts.gate_up_proj": "model-00016-of-00018.safetensors",
+    "model.layers.22.mlp.experts.gate_up_proj_bias": "model-00016-of-00018.safetensors",
+    "model.layers.22.mlp.router.bias": "model-00016-of-00018.safetensors",
+    "model.layers.22.mlp.router.weight": "model-00016-of-00018.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "model-00016-of-00018.safetensors",
+    "model.layers.22.self_attn.k_proj.bias": "model-00016-of-00018.safetensors",
+    "model.layers.22.self_attn.k_proj.weight": "model-00016-of-00018.safetensors",
+    "model.layers.22.self_attn.o_proj.bias": "model-00016-of-00018.safetensors",
+    "model.layers.22.self_attn.o_proj.weight": "model-00016-of-00018.safetensors",
+    "model.layers.22.self_attn.q_proj.bias": "model-00016-of-00018.safetensors",
+    "model.layers.22.self_attn.q_proj.weight": "model-00016-of-00018.safetensors",
+    "model.layers.22.self_attn.sinks": "model-00016-of-00018.safetensors",
+    "model.layers.22.self_attn.v_proj.bias": "model-00016-of-00018.safetensors",
+    "model.layers.22.self_attn.v_proj.weight": "model-00016-of-00018.safetensors",
+    "model.layers.23.input_layernorm.weight": "model-00017-of-00018.safetensors",
+    "model.layers.23.mlp.experts.down_proj": "model-00017-of-00018.safetensors",
+    "model.layers.23.mlp.experts.down_proj_bias": "model-00017-of-00018.safetensors",
+    "model.layers.23.mlp.experts.gate_up_proj": "model-00017-of-00018.safetensors",
+    "model.layers.23.mlp.experts.gate_up_proj_bias": "model-00017-of-00018.safetensors",
+    "model.layers.23.mlp.router.bias": "model-00016-of-00018.safetensors",
+    "model.layers.23.mlp.router.weight": "model-00016-of-00018.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "model-00017-of-00018.safetensors",
+    "model.layers.23.self_attn.k_proj.bias": "model-00016-of-00018.safetensors",
+    "model.layers.23.self_attn.k_proj.weight": "model-00016-of-00018.safetensors",
+    "model.layers.23.self_attn.o_proj.bias": "model-00016-of-00018.safetensors",
+    "model.layers.23.self_attn.o_proj.weight": "model-00016-of-00018.safetensors",
+    "model.layers.23.self_attn.q_proj.bias": "model-00016-of-00018.safetensors",
+    "model.layers.23.self_attn.q_proj.weight": "model-00016-of-00018.safetensors",
+    "model.layers.23.self_attn.sinks": "model-00016-of-00018.safetensors",
+    "model.layers.23.self_attn.v_proj.bias": "model-00016-of-00018.safetensors",
+    "model.layers.23.self_attn.v_proj.weight": "model-00016-of-00018.safetensors",
+    "model.layers.3.input_layernorm.weight": "model-00004-of-00018.safetensors",
+    "model.layers.3.mlp.experts.down_proj": "model-00004-of-00018.safetensors",
+    "model.layers.3.mlp.experts.down_proj_bias": "model-00004-of-00018.safetensors",
+    "model.layers.3.mlp.experts.gate_up_proj": "model-00003-of-00018.safetensors",
+    "model.layers.3.mlp.experts.gate_up_proj_bias": "model-00003-of-00018.safetensors",
+    "model.layers.3.mlp.router.bias": "model-00003-of-00018.safetensors",
+    "model.layers.3.mlp.router.weight": "model-00003-of-00018.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "model-00004-of-00018.safetensors",
+    "model.layers.3.self_attn.k_proj.bias": "model-00003-of-00018.safetensors",
+    "model.layers.3.self_attn.k_proj.weight": "model-00003-of-00018.safetensors",
+    "model.layers.3.self_attn.o_proj.bias": "model-00003-of-00018.safetensors",
+    "model.layers.3.self_attn.o_proj.weight": "model-00003-of-00018.safetensors",
+    "model.layers.3.self_attn.q_proj.bias": "model-00003-of-00018.safetensors",
+    "model.layers.3.self_attn.q_proj.weight": "model-00003-of-00018.safetensors",
+    "model.layers.3.self_attn.sinks": "model-00003-of-00018.safetensors",
+    "model.layers.3.self_attn.v_proj.bias": "model-00003-of-00018.safetensors",
+    "model.layers.3.self_attn.v_proj.weight": "model-00003-of-00018.safetensors",
+    "model.layers.4.input_layernorm.weight": "model-00004-of-00018.safetensors",
+    "model.layers.4.mlp.experts.down_proj": "model-00004-of-00018.safetensors",
+    "model.layers.4.mlp.experts.down_proj_bias": "model-00004-of-00018.safetensors",
+    "model.layers.4.mlp.experts.gate_up_proj": "model-00004-of-00018.safetensors",
+    "model.layers.4.mlp.experts.gate_up_proj_bias": "model-00004-of-00018.safetensors",
+    "model.layers.4.mlp.router.bias": "model-00004-of-00018.safetensors",
+    "model.layers.4.mlp.router.weight": "model-00004-of-00018.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "model-00004-of-00018.safetensors",
+    "model.layers.4.self_attn.k_proj.bias": "model-00004-of-00018.safetensors",
+    "model.layers.4.self_attn.k_proj.weight": "model-00004-of-00018.safetensors",
+    "model.layers.4.self_attn.o_proj.bias": "model-00004-of-00018.safetensors",
+    "model.layers.4.self_attn.o_proj.weight": "model-00004-of-00018.safetensors",
+    "model.layers.4.self_attn.q_proj.bias": "model-00004-of-00018.safetensors",
+    "model.layers.4.self_attn.q_proj.weight": "model-00004-of-00018.safetensors",
+    "model.layers.4.self_attn.sinks": "model-00004-of-00018.safetensors",
+    "model.layers.4.self_attn.v_proj.bias": "model-00004-of-00018.safetensors",
+    "model.layers.4.self_attn.v_proj.weight": "model-00004-of-00018.safetensors",
+    "model.layers.5.input_layernorm.weight": "model-00005-of-00018.safetensors",
+    "model.layers.5.mlp.experts.down_proj": "model-00005-of-00018.safetensors",
+    "model.layers.5.mlp.experts.down_proj_bias": "model-00005-of-00018.safetensors",
+    "model.layers.5.mlp.experts.gate_up_proj": "model-00005-of-00018.safetensors",
+    "model.layers.5.mlp.experts.gate_up_proj_bias": "model-00005-of-00018.safetensors",
+    "model.layers.5.mlp.router.bias": "model-00004-of-00018.safetensors",
+    "model.layers.5.mlp.router.weight": "model-00004-of-00018.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "model-00005-of-00018.safetensors",
+    "model.layers.5.self_attn.k_proj.bias": "model-00004-of-00018.safetensors",
+    "model.layers.5.self_attn.k_proj.weight": "model-00004-of-00018.safetensors",
+    "model.layers.5.self_attn.o_proj.bias": "model-00004-of-00018.safetensors",
+    "model.layers.5.self_attn.o_proj.weight": "model-00004-of-00018.safetensors",
+    "model.layers.5.self_attn.q_proj.bias": "model-00004-of-00018.safetensors",
+    "model.layers.5.self_attn.q_proj.weight": "model-00004-of-00018.safetensors",
+    "model.layers.5.self_attn.sinks": "model-00004-of-00018.safetensors",
+    "model.layers.5.self_attn.v_proj.bias": "model-00004-of-00018.safetensors",
+    "model.layers.5.self_attn.v_proj.weight": "model-00004-of-00018.safetensors",
+    "model.layers.6.input_layernorm.weight": "model-00006-of-00018.safetensors",
+    "model.layers.6.mlp.experts.down_proj": "model-00006-of-00018.safetensors",
+    "model.layers.6.mlp.experts.down_proj_bias": "model-00006-of-00018.safetensors",
+    "model.layers.6.mlp.experts.gate_up_proj": "model-00005-of-00018.safetensors",
+    "model.layers.6.mlp.experts.gate_up_proj_bias": "model-00005-of-00018.safetensors",
+    "model.layers.6.mlp.router.bias": "model-00005-of-00018.safetensors",
+    "model.layers.6.mlp.router.weight": "model-00005-of-00018.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "model-00006-of-00018.safetensors",
+    "model.layers.6.self_attn.k_proj.bias": "model-00005-of-00018.safetensors",
+    "model.layers.6.self_attn.k_proj.weight": "model-00005-of-00018.safetensors",
+    "model.layers.6.self_attn.o_proj.bias": "model-00005-of-00018.safetensors",
+    "model.layers.6.self_attn.o_proj.weight": "model-00005-of-00018.safetensors",
+    "model.layers.6.self_attn.q_proj.bias": "model-00005-of-00018.safetensors",
+    "model.layers.6.self_attn.q_proj.weight": "model-00005-of-00018.safetensors",
+    "model.layers.6.self_attn.sinks": "model-00005-of-00018.safetensors",
+    "model.layers.6.self_attn.v_proj.bias": "model-00005-of-00018.safetensors",
+    "model.layers.6.self_attn.v_proj.weight": "model-00005-of-00018.safetensors",
+    "model.layers.7.input_layernorm.weight": "model-00006-of-00018.safetensors",
+    "model.layers.7.mlp.experts.down_proj": "model-00006-of-00018.safetensors",
+    "model.layers.7.mlp.experts.down_proj_bias": "model-00006-of-00018.safetensors",
+    "model.layers.7.mlp.experts.gate_up_proj": "model-00006-of-00018.safetensors",
+    "model.layers.7.mlp.experts.gate_up_proj_bias": "model-00006-of-00018.safetensors",
+    "model.layers.7.mlp.router.bias": "model-00006-of-00018.safetensors",
+    "model.layers.7.mlp.router.weight": "model-00006-of-00018.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "model-00006-of-00018.safetensors",
+    "model.layers.7.self_attn.k_proj.bias": "model-00006-of-00018.safetensors",
+    "model.layers.7.self_attn.k_proj.weight": "model-00006-of-00018.safetensors",
+    "model.layers.7.self_attn.o_proj.bias": "model-00006-of-00018.safetensors",
+    "model.layers.7.self_attn.o_proj.weight": "model-00006-of-00018.safetensors",
+    "model.layers.7.self_attn.q_proj.bias": "model-00006-of-00018.safetensors",
+    "model.layers.7.self_attn.q_proj.weight": "model-00006-of-00018.safetensors",
+    "model.layers.7.self_attn.sinks": "model-00006-of-00018.safetensors",
+    "model.layers.7.self_attn.v_proj.bias": "model-00006-of-00018.safetensors",
+    "model.layers.7.self_attn.v_proj.weight": "model-00006-of-00018.safetensors",
+    "model.layers.8.input_layernorm.weight": "model-00007-of-00018.safetensors",
+    "model.layers.8.mlp.experts.down_proj": "model-00007-of-00018.safetensors",
+    "model.layers.8.mlp.experts.down_proj_bias": "model-00007-of-00018.safetensors",
+    "model.layers.8.mlp.experts.gate_up_proj": "model-00007-of-00018.safetensors",
+    "model.layers.8.mlp.experts.gate_up_proj_bias": "model-00007-of-00018.safetensors",
+    "model.layers.8.mlp.router.bias": "model-00006-of-00018.safetensors",
+    "model.layers.8.mlp.router.weight": "model-00006-of-00018.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "model-00007-of-00018.safetensors",
+    "model.layers.8.self_attn.k_proj.bias": "model-00006-of-00018.safetensors",
+    "model.layers.8.self_attn.k_proj.weight": "model-00006-of-00018.safetensors",
+    "model.layers.8.self_attn.o_proj.bias": "model-00006-of-00018.safetensors",
+    "model.layers.8.self_attn.o_proj.weight": "model-00006-of-00018.safetensors",
+    "model.layers.8.self_attn.q_proj.bias": "model-00006-of-00018.safetensors",
+    "model.layers.8.self_attn.q_proj.weight": "model-00006-of-00018.safetensors",
+    "model.layers.8.self_attn.sinks": "model-00006-of-00018.safetensors",
+    "model.layers.8.self_attn.v_proj.bias": "model-00006-of-00018.safetensors",
+    "model.layers.8.self_attn.v_proj.weight": "model-00006-of-00018.safetensors",
+    "model.layers.9.input_layernorm.weight": "model-00008-of-00018.safetensors",
+    "model.layers.9.mlp.experts.down_proj": "model-00008-of-00018.safetensors",
+    "model.layers.9.mlp.experts.down_proj_bias": "model-00008-of-00018.safetensors",
+    "model.layers.9.mlp.experts.gate_up_proj": "model-00007-of-00018.safetensors",
+    "model.layers.9.mlp.experts.gate_up_proj_bias": "model-00007-of-00018.safetensors",
+    "model.layers.9.mlp.router.bias": "model-00007-of-00018.safetensors",
+    "model.layers.9.mlp.router.weight": "model-00007-of-00018.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "model-00008-of-00018.safetensors",
+    "model.layers.9.self_attn.k_proj.bias": "model-00007-of-00018.safetensors",
+    "model.layers.9.self_attn.k_proj.weight": "model-00007-of-00018.safetensors",
+    "model.layers.9.self_attn.o_proj.bias": "model-00007-of-00018.safetensors",
+    "model.layers.9.self_attn.o_proj.weight": "model-00007-of-00018.safetensors",
+    "model.layers.9.self_attn.q_proj.bias": "model-00007-of-00018.safetensors",
+    "model.layers.9.self_attn.q_proj.weight": "model-00007-of-00018.safetensors",
+    "model.layers.9.self_attn.sinks": "model-00007-of-00018.safetensors",
+    "model.layers.9.self_attn.v_proj.bias": "model-00007-of-00018.safetensors",
+    "model.layers.9.self_attn.v_proj.weight": "model-00007-of-00018.safetensors",
+    "model.norm.weight": "model-00017-of-00018.safetensors"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,23 @@
+{
+  "bos_token": {
+    "content": "<|startoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "<|return|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0614fe83cadab421296e664e1f48f4261fa8fef6e03e63bb75c20f38e37d07d3
+size 27868174

tokenizer_config.json

@@ -0,0 +1,183 @@
+{
+  "added_tokens_decoder": {
+    "199998": {
+      "content": "<|startoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "199999": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200000": {
+      "content": "<|reserved_200000|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200001": {
+      "content": "<|reserved_200001|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200002": {
+      "content": "<|return|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200003": {
+      "content": "<|constrain|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200004": {
+      "content": "<|reserved_200004|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200005": {
+      "content": "<|channel|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200006": {
+      "content": "<|start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200007": {
+      "content": "<|end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200008": {
+      "content": "<|message|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200009": {
+      "content": "<|reserved_200009|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200010": {
+      "content": "<|reserved_200010|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200011": {
+      "content": "<|reserved_200011|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200012": {
+      "content": "<|call|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200013": {
+      "content": "<|reserved_200013|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200014": {
+      "content": "<|reserved_200014|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200015": {
+      "content": "<|reserved_200015|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200016": {
+      "content": "<|reserved_200016|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200017": {
+      "content": "<|reserved_200017|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "200018": {
+      "content": "<|endofprompt|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<|startoftext|>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|return|>",
+  "extra_special_tokens": {},
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|endoftext|>",
+  "tokenizer_class": "PreTrainedTokenizerFast"
+}