Upload README.md with huggingface_hub

README.md

@@ -1,199 +1,204 @@
 ---
+license: mit
+license_link: https://huggingface.co/microsoft/phi-4/resolve/main/LICENSE
+language:
+- en
+pipeline_tag: text-generation
+tags:
+- phi
+- nlp
+- math
+- code
+- chat
+- conversational
+- heretic
+- uncensored
+- decensored
+- abliterated
+inference:
+  parameters:
+    temperature: 0
+widget:
+- messages:
+  - role: user
+    content: How should I explain the Internet?
 library_name: transformers
-tags: []
 ---
+# This is a decensored version of [microsoft/phi-4](https://huggingface.co/microsoft/phi-4), made using [Heretic](https://github.com/p-e-w/heretic) v1.0.0
 
-# Model Card for Model ID
+## Abliteration parameters
 
-<!-- Provide a quick summary of what the model is/does. -->
+| Parameter | Value |
+| :-------- | :---: |
+| **direction_index** | 15.75 |
+| **attn.o_proj.max_weight** | 1.48 |
+| **attn.o_proj.max_weight_position** | 25.38 |
+| **attn.o_proj.min_weight** | 0.19 |
+| **attn.o_proj.min_weight_distance** | 11.85 |
+| **mlp.down_proj.max_weight** | 1.45 |
+| **mlp.down_proj.max_weight_position** | 26.40 |
+| **mlp.down_proj.min_weight** | 1.43 |
+| **mlp.down_proj.min_weight_distance** | 23.33 |
 
+## Performance
 
+| Metric | This model | Original model ([microsoft/phi-4](https://huggingface.co/microsoft/phi-4)) |
+| :----- | :--------: | :---------------------------: |
+| **KL divergence** | 0.09 | 0 *(by definition)* |
+| **Refusals** | 41/100 | 100/100 |
 
-## Model Details
+-----
 
-### Model Description
 
-<!-- Provide a longer summary of what this model is. -->
+# Phi-4 Model Card 
 
-This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
+[Phi-4 Technical Report](https://arxiv.org/pdf/2412.08905)
 
-- **Developed by:** [More Information Needed]
-- **Funded by [optional]:** [More Information Needed]
-- **Shared by [optional]:** [More Information Needed]
-- **Model type:** [More Information Needed]
-- **Language(s) (NLP):** [More Information Needed]
-- **License:** [More Information Needed]
-- **Finetuned from model [optional]:** [More Information Needed]
+## Model Summary 
 
-### Model Sources [optional]
+|                         |                                                                               |     
+|-------------------------|-------------------------------------------------------------------------------|
+| **Developers**          | Microsoft Research                                                            |
+| **Description**         | `phi-4` is a state-of-the-art open model built upon a blend of synthetic datasets, data from filtered public domain websites, and acquired academic books and Q&A datasets. The goal of this approach was to ensure that small capable models were trained with data focused on high quality and advanced reasoning.<br><br>`phi-4` underwent a rigorous enhancement and alignment process, incorporating both supervised fine-tuning and direct preference optimization to ensure precise instruction adherence and robust safety measures                |
+| **Architecture**        | 14B parameters, dense decoder-only Transformer model                          |
+| **Inputs**              | Text, best suited for prompts in the chat format                              |
+| **Context length**      | 16K tokens                                                                    |
+| **GPUs**                | 1920 H100-80G                                                                 |
+| **Training time**       | 21 days                                                                       |
+| **Training data**       | 9.8T tokens                                                                   |
+| **Outputs**             | Generated text in response to input                                           |
+| **Dates**               | October 2024 – November 2024                                                  |
+| **Status**              | Static model trained on an offline dataset with cutoff dates of June 2024 and earlier for publicly available data                                                                               |
+| **Release date**        | December 12, 2024                                                             |
+| **License**             | MIT                                                                         |
 
-<!-- Provide the basic links for the model. -->
+## Intended Use 
 
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
+|                               |                                                                         |
+|-------------------------------|-------------------------------------------------------------------------|
+| **Primary Use Cases**         | Our model is designed to accelerate research on language models, for use as a building block for generative AI powered features. It provides uses for general purpose AI systems and applications (primarily in English) which require:<br><br>1. Memory/compute constrained environments.<br>2. Latency bound scenarios.<br>3. Reasoning and logic.                                                                       |
+| **Out-of-Scope Use Cases**    | Our models is not specifically designed or evaluated for all downstream purposes, thus:<br><br>1. Developers should consider common limitations of language models as they select use cases, and evaluate and mitigate for accuracy, safety, and fairness before using within a specific downstream use case, particularly for high-risk scenarios.<br>2. Developers should be aware of and adhere to applicable laws or regulations (including privacy, trade compliance laws, etc.) that are relevant to their use case, including the model’s focus on English.<br>3. Nothing contained in this Model Card should be interpreted as or deemed a restriction or modification to the license the model is released under.                                                              |
 
-## Uses
+## Data Overview 
 
-<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+### Training Datasets 
 
-### Direct Use
+Our training data is an extension of the data used for Phi-3 and includes a wide variety of sources from:
 
-<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+1. Publicly available documents filtered rigorously for quality, selected high-quality educational data, and code.
 
-[More Information Needed]
+2. Newly created synthetic, “textbook-like” data for the purpose of teaching math, coding, common sense reasoning, general knowledge of the world (science, daily activities, theory of mind, etc.).
 
-### Downstream Use [optional]
+3. Acquired academic books and Q&A datasets.
 
-<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+4. High quality chat format supervised data covering various topics to reflect human preferences on different aspects such as instruct-following, truthfulness, honesty and helpfulness.
 
-[More Information Needed]
+Multilingual data constitutes about 8% of our overall data. We are focusing on the quality of data that could potentially improve the reasoning ability for the model, and we filter the publicly available documents to contain the correct level of knowledge.
 
-### Out-of-Scope Use
+#### Benchmark datasets 
 
-<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+We evaluated `phi-4` using [OpenAI’s SimpleEval](https://github.com/openai/simple-evals) and our own internal benchmarks to understand the model’s capabilities, more specifically: 
 
-[More Information Needed]
+* **MMLU:** Popular aggregated dataset for multitask language understanding.
 
-## Bias, Risks, and Limitations
+* **MATH:** Challenging competition math problems.
 
-<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+* **GPQA:** Complex, graduate-level science questions.
 
-[More Information Needed]
+* **DROP:** Complex comprehension and reasoning.
 
-### Recommendations
+* **MGSM:** Multi-lingual grade-school math.
 
-<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+* **HumanEval:** Functional code generation.
 
-Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+* **SimpleQA:** Factual responses.
 
-## How to Get Started with the Model
+## Safety 
 
-Use the code below to get started with the model.
+### Approach 
 
-[More Information Needed]
+`phi-4` has adopted a robust safety post-training approach. This approach leverages a variety of both open-source and in-house generated synthetic datasets. The overall technique employed to do the safety alignment is a combination of SFT (Supervised Fine-Tuning) and iterative DPO (Direct Preference Optimization), including publicly available datasets focusing on helpfulness and harmlessness as well as various questions and answers targeted to multiple safety categories. 
 
-## Training Details
+### Safety Evaluation and Red-Teaming 
 
-### Training Data
+Prior to release, `phi-4` followed a multi-faceted evaluation approach. Quantitative evaluation was conducted with multiple open-source safety benchmarks and in-house tools utilizing adversarial conversation simulation. For qualitative safety evaluation, we collaborated with the independent AI Red Team (AIRT) at Microsoft to assess safety risks posed by `phi-4` in both average and adversarial user scenarios. In the average user scenario, AIRT emulated typical single-turn and multi-turn interactions to identify potentially risky behaviors. The adversarial user scenario tested a wide range of techniques aimed at intentionally subverting the model’s safety training including jailbreaks, encoding-based attacks, multi-turn attacks, and adversarial suffix attacks.   
 
-<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+Please refer to the technical report for more details on safety alignment. 
 
-[More Information Needed]
+## Model Quality
 
-### Training Procedure
+To understand the capabilities, we compare `phi-4` with a set of models over OpenAI’s SimpleEval benchmark. 
 
-<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+At the high-level overview of the model quality on representative benchmarks. For the table below, higher numbers indicate better performance: 
 
-#### Preprocessing [optional]
+| **Category**                 | **Benchmark** | **phi-4** (14B) | **phi-3** (14B) | **Qwen 2.5** (14B instruct) | **GPT-4o-mini** | **Llama-3.3** (70B instruct) | **Qwen 2.5** (72B instruct) | **GPT-4o** |
+|------------------------------|---------------|-----------|-----------------|----------------------|----------------------|--------------------|-------------------|-----------------|
+| Popular Aggregated Benchmark | MMLU          | 84.8      | 77.9            | 79.9                 | 81.8                 | 86.3               | 85.3              | **88.1**            |
+| Science                      | GPQA          | **56.1**      | 31.2            | 42.9                 | 40.9                 | 49.1               | 49.0              | 50.6            |
+| Math                         | MGSM<br>MATH  | 80.6<br>**80.4** | 53.5<br>44.6 | 79.6<br>75.6 | 86.5<br>73.0 | 89.1<br>66.3* | 87.3<br>80.0              | **90.4**<br>74.6            |
+| Code Generation              | HumanEval     | 82.6      | 67.8            | 72.1                 | 86.2                 | 78.9*               | 80.4              | **90.6**            |
+| Factual Knowledge            | SimpleQA      | 3.0       | 7.6            | 5.4                 | 9.9                  | 20.9               | 10.2              | **39.4**             |
+| Reasoning                    | DROP          | 75.5      | 68.3            | 85.5                 | 79.3                 | **90.2**               | 76.7              | 80.9            |
 
-[More Information Needed]
+\* These scores are lower than those reported by Meta, perhaps because simple-evals has a strict formatting requirement that Llama models have particular trouble following. We use the simple-evals framework because it is reproducible, but Meta reports 77 for MATH and 88 for HumanEval on Llama-3.3-70B.
 
+## Usage
 
-#### Training Hyperparameters
+### Input Formats
 
-- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+Given the nature of the training data, `phi-4` is best suited for prompts using the chat format as follows: 
 
-#### Speeds, Sizes, Times [optional]
+```bash
+<|im_start|>system<|im_sep|>
+You are a medieval knight and must provide explanations to modern people.<|im_end|>
+<|im_start|>user<|im_sep|>
+How should I explain the Internet?<|im_end|>
+<|im_start|>assistant<|im_sep|>
+```
 
-<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+### With `transformers`
 
-[More Information Needed]
+```python
+import transformers
 
-## Evaluation
+pipeline = transformers.pipeline(
+    "text-generation",
+    model="microsoft/phi-4",
+    model_kwargs={"torch_dtype": "auto"},
+    device_map="auto",
+)
 
-<!-- This section describes the evaluation protocols and provides the results. -->
+messages = [
+    {"role": "system", "content": "You are a medieval knight and must provide explanations to modern people."},
+    {"role": "user", "content": "How should I explain the Internet?"},
+]
 
-### Testing Data, Factors & Metrics
+outputs = pipeline(messages, max_new_tokens=128)
+print(outputs[0]["generated_text"][-1])
+```
 
-#### Testing Data
+## Responsible AI Considerations
 
-<!-- This should link to a Dataset Card if possible. -->
+Like other language models, `phi-4` can potentially behave in ways that are unfair, unreliable, or offensive. Some of the limiting behaviors to be aware of include:   
 
-[More Information Needed]
+* **Quality of Service:** The model is trained primarily on English text. Languages other than English will experience worse performance. English language varieties with less representation in the training data might experience worse performance than standard American English. `phi-4` is not intended to support multilingual use. 
 
-#### Factors
+* **Representation of Harms & Perpetuation of Stereotypes:** These models can over- or under-represent groups of people, erase representation of some groups, or reinforce demeaning or negative stereotypes. Despite safety post-training, these limitations may still be present due to differing levels of representation of different groups or prevalence of examples of negative stereotypes in training data that reflect real-world patterns and societal biases.  
 
-<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+* **Inappropriate or Offensive Content:** These models may produce other types of inappropriate or offensive content, which may make it inappropriate to deploy for sensitive contexts without additional mitigations that are specific to the use case.  
 
-[More Information Needed]
+* **Information Reliability:** Language models can generate nonsensical content or fabricate content that might sound reasonable but is inaccurate or outdated.   
 
-#### Metrics
+* **Limited Scope for Code:** Majority of `phi-4` training data is based in Python and uses common packages such as `typing`, `math`, `random`, `collections`, `datetime`, `itertools`. If the model generates Python scripts that utilize other packages or scripts in other languages, we strongly recommend users manually verify all API uses.  
 
-<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+Developers should apply responsible AI best practices and are responsible for ensuring that a specific use case complies with relevant laws and regulations (e.g. privacy, trade, etc.). Using safety services like [Azure AI Content Safety](https://azure.microsoft.com/en-us/products/ai-services/ai-content-safety) that have advanced guardrails is highly recommended. Important areas for consideration include:
 
-[More Information Needed]
+* **Allocation:** Models may not be suitable for scenarios that could have consequential impact on legal status or the allocation of resources or life opportunities (ex: housing, employment, credit, etc.) without further assessments and additional debiasing techniques. 
 
-### Results
+* **High-Risk Scenarios:** Developers should assess suitability of using models in high-risk scenarios where unfair, unreliable or offensive outputs might be extremely costly or lead to harm. This includes providing advice in sensitive or expert domains where accuracy and reliability are critical (ex: legal or health advice). Additional safeguards should be implemented at the application level according to the deployment context.  
 
-[More Information Needed]
+* **Misinformation:** Models may produce inaccurate information. Developers should follow transparency best practices and inform end-users they are interacting with an AI system. At the application level, developers can build feedback mechanisms and pipelines to ground responses in use-case specific, contextual information, a technique known as Retrieval Augmented Generation (RAG).    
 
-#### Summary
+* **Generation of Harmful Content:** Developers should assess outputs for their context and use available safety classifiers or custom solutions appropriate for their use case.  
 
-
-
-## Model Examination [optional]
-
-<!-- Relevant interpretability work for the model goes here -->
-
-[More Information Needed]
-
-## Environmental Impact
-
-<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
-
-Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
-
-- **Hardware Type:** [More Information Needed]
-- **Hours used:** [More Information Needed]
-- **Cloud Provider:** [More Information Needed]
-- **Compute Region:** [More Information Needed]
-- **Carbon Emitted:** [More Information Needed]
-
-## Technical Specifications [optional]
-
-### Model Architecture and Objective
-
-[More Information Needed]
-
-### Compute Infrastructure
-
-[More Information Needed]
-
-#### Hardware
-
-[More Information Needed]
-
-#### Software
-
-[More Information Needed]
-
-## Citation [optional]
-
-<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
-
-**BibTeX:**
-
-[More Information Needed]
-
-**APA:**
-
-[More Information Needed]
-
-## Glossary [optional]
-
-<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
-
-[More Information Needed]
-
-## More Information [optional]
-
-[More Information Needed]
-
-## Model Card Authors [optional]
-
-[More Information Needed]
-
-## Model Card Contact
-
-[More Information Needed]
+* **Misuse:** Other forms of misuse such as fraud, spam, or malware production may be possible, and developers should ensure that their applications do not violate applicable laws and regulations.