File size: 18,238 Bytes
8942a0b 349da96 8942a0b 1ecb710 62066f8 1ecb710 8942a0b 349da96 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 |
---
base_model: unsloth/phi-3.5-mini-instruct-bnb-4bit
language:
- en
license: apache-2.0
tags:
- text-generation-inference
- transformers
- unsloth
- llama
- trl
pipeline_tag: question-answering
---
## Model Summary
Athene-Phi-3.5-mini-instruct is a lightweight, state-of-the-art open model built upon datasets used for Phi-3 - synthetic data and filtered publicly available websites - with a focus on very high-quality, reasoning dense data. The model belongs to the Phi-3 model family and supports 128K token context length. The model underwent a rigorous enhancement process, incorporating both supervised fine-tuning, proximal policy optimization, and direct preference optimization to ensure precise instruction adherence and robust safety measures.
🏡 [Phi-3 Portal](https://azure.microsoft.com/en-us/products/phi-3) <br>
📰 [Phi-3 Microsoft Blog](https://aka.ms/phi3.5-techblog) <br>
📖 [Phi-3 Technical Report](https://arxiv.org/abs/2404.14219) <br>
👩🍳 [Phi-3 Cookbook](https://github.com/microsoft/Phi-3CookBook) <br>
🖥️ [Try It](https://aka.ms/try-phi3.5mini) <br>
**Phi-3.5**: [[mini-instruct]](https://huggingface.co/microsoft/Phi-3.5-mini-instruct); [[MoE-instruct]](https://huggingface.co/microsoft/Phi-3.5-MoE-instruct) ; [[vision-instruct]](https://huggingface.co/microsoft/Phi-3.5-vision-instruct)
## Intended Uses
### Primary Use Cases
The model is intended for commercial and research use in multiple languages. The model provides uses for general purpose AI systems and applications which require:
1) Memory/compute constrained environments
2) Latency bound scenarios
3) Strong reasoning (especially code, math and logic)
Our model is designed to accelerate research on language and multimodal models, for use as a building block for generative AI powered features.
### Use Case Considerations
Our models are not specifically designed or evaluated for all downstream purposes. Developers should consider common limitations of language models as they select use cases, and evaluate and mitigate for accuracy, safety, and fariness before using within a specific downstream use case, particularly for high risk scenarios. 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.
***Nothing contained in this Model Card should be interpreted as or deemed a restriction or modification to the license the model is released under.***
## Release Notes
This is an update over the June 2024 instruction-tuned Phi-3 Mini release based on valuable user feedback. The model used additional post-training data leading to substantial gains on multilingual, multi-turn conversation quality, and reasoning capability. We believe most use cases will benefit from this release, but we encourage users to test in their particular AI applications. We appreciate the enthusiastic adoption of the Phi-3 model family, and continue to welcome all feedback from the community.
### Multilingual
The table below highlights multilingual capability of the Phi-3.5 Mini on multilingual MMLU, MEGA, and multilingual MMLU-pro datasets. Overall, we observed that even with just 3.8B active parameters, the model is competitive on multilingual tasks in comparison to other models with a much bigger active parameters.
| Benchmark | Phi-3.5 Mini-Ins | Phi-3.1-Mini-128K-Ins | Mistral-7B-Instruct-v0.3 | Mistral-Nemo-12B-Ins-2407 | Llama-3.1-8B-Ins | Gemma-2-9B-Ins | Gemini 1.5 Flash | GPT-4o-mini-2024-07-18 (Chat) |
|----------------------------|------------------|-----------------------|--------------------------|---------------------------|------------------|----------------|------------------|-------------------------------|
| Multilingual MMLU | 55.4 | 51.08 | 47.4 | 58.9 | 56.2 | 63.8 | 77.2 | 72.9 |
| Multilingual MMLU-Pro | 30.9 | 30.21 | 15.0 | 34.0 | 21.4 | 43.0 | 57.9 | 53.2 |
| MGSM | 47.9 | 41.56 | 31.8 | 63.3 | 56.7 | 75.1 | 75.8 | 81.7 |
| MEGA MLQA | 61.7 | 55.5 | 43.9 | 61.2 | 45.2 | 54.4 | 61.6 | 70.0 |
| MEGA TyDi QA | 62.2 | 55.9 | 54.0 | 63.7 | 54.5 | 65.6 | 63.6 | 81.8 |
| MEGA UDPOS | 46.5 | 48.1 | 57.2 | 58.2 | 54.1 | 56.6 | 62.4 | 66.0 |
| MEGA XCOPA | 63.1 | 62.4 | 58.8 | 10.8 | 21.1 | 31.2 | 95.0 | 90.3 |
| MEGA XStoryCloze | 73.5 | 73.6 | 75.5 | 92.3 | 71.0 | 87.0 | 20.7 | 96.6 |
| **Average** | **55.2** | **52.3** | **47.9** | **55.3** | **47.5** | **59.6** | **64.3** | **76.6** |
The table below shows Multilingual MMLU scores in some of the supported languages. For more multi-lingual benchmarks and details, see [Appendix A](#appendix-a).
| Benchmark | Phi-3.5 Mini-Ins | Phi-3.1-Mini-128K-Ins | Mistral-7B-Instruct-v0.3 | Mistral-Nemo-12B-Ins-2407 | Llama-3.1-8B-Ins | Gemma-2-9B-Ins | Gemini 1.5 Flash | GPT-4o-mini-2024-07-18 (Chat) |
|-----------|------------------|-----------------------|--------------------------|---------------------------|------------------|----------------|------------------|-------------------------------|
| Arabic | 44.2 | 35.4 | 33.7 | 45.3 | 49.1 | 56.3 | 73.6 | 67.1 |
| Chinese | 52.6 | 46.9 | 45.9 | 58.2 | 54.4 | 62.7 | 66.7 | 70.8 |
| Dutch | 57.7 | 48.0 | 51.3 | 60.1 | 55.9 | 66.7 | 80.6 | 74.2 |
| French | 61.1 | 61.7 | 53.0 | 63.8 | 62.8 | 67.0 | 82.9 | 75.6 |
| German | 62.4 | 61.3 | 50.1 | 64.5 | 59.9 | 65.7 | 79.5 | 74.3 |
| Italian | 62.8 | 63.1 | 52.5 | 64.1 | 55.9 | 65.7 | 82.6 | 75.9 |
| Russian | 50.4 | 45.3 | 48.9 | 59.0 | 57.4 | 63.2 | 78.7 | 72.6 |
| Spanish | 62.6 | 61.3 | 53.9 | 64.3 | 62.6 | 66.0 | 80.0 | 75.5 |
| Ukrainian | 45.2 | 36.7 | 46.9 | 56.6 | 52.9 | 62.0 | 77.4 | 72.6 |
### Long Context
Phi-3.5-mini supports 128K context length, therefore the model is capable of several long context tasks including long document/meeting summarization, long document QA, long document information retrieval. We see that Phi-3.5-mini is clearly better than Gemma-2 family which only supports 8K context length. Phi-3.5-mini is competitive with other much larger open-weight models such as Llama-3.1-8B-instruct, Mistral-7B-instruct-v0.3, and Mistral-Nemo-12B-instruct-2407.
| Benchmark | Phi-3.5-mini-instruct | Llama-3.1-8B-instruct | Mistral-7B-instruct-v0.3 | Mistral-Nemo-12B-instruct-2407 | Gemini-1.5-Flash | GPT-4o-mini-2024-07-18 (Chat) |
|--|--|--|--|--|--|--|
| GovReport | 25.9 | 25.1 | 26.0 | 25.6 | 27.8 | 24.8 |
| QMSum | 21.3 | 21.6 | 21.3 | 22.1 | 24.0 | 21.7 |
| Qasper | 41.9 | 37.2 | 31.4 | 30.7 | 43.5 | 39.8 |
| SQuALITY | 25.3 | 26.2 | 25.9 | 25.8 | 23.5 | 23.8 |
| SummScreenFD | 16.0 | 17.6 | 17.5 | 18.2 | 16.3 | 17.0 |
| **Average** | **26.1** | **25.5** | **24.4** | **24.5** | **27.0** | **25.4** |
RULER: a retrieval-based benchmark for long context understanding
| Model | 4K | 8K | 16K | 32K | 64K | 128K | Average |
|--|--|--|--|--|--|--|--|
| **Phi-3.5-mini-instruct** | 94.3 | 91.1 | 90.7 | 87.1 | 78.0 | 63.6 | **84.1** |
| **Llama-3.1-8B-instruct** | 95.5 | 93.8 | 91.6 | 87.4 | 84.7 | 77.0 | **88.3** |
| **Mistral-Nemo-12B-instruct-2407** | 87.8 | 87.2 | 87.7 | 69.0 | 46.8 | 19.0 | **66.2** |
RepoQA: a benchmark for long context code understanding
| Model | Python | C++ | Rust | Java | TypeScript | Average |
|--|--|--|--|--|--|--|
| **Phi-3.5-mini-instruct** | 86 | 67 | 73 | 77 | 82 | **77** |
| **Llama-3.1-8B-instruct** | 80 | 65 | 73 | 76 | 63 | **71** |
| **Mistral-7B-instruct-v0.3** | 61 | 57 | 51 | 61 | 80 | **62** |
## Usage
### Requirements
Phi-3 family has been integrated in the `4.43.0` version of `transformers`. The current `transformers` version can be verified with: `pip list | grep transformers`.
Examples of required packages:
```
flash_attn==2.5.8
torch==2.3.1
accelerate==0.31.0
transformers==4.43.0
```
Phi-3.5-mini-instruct is also available in [Azure AI Studio](https://aka.ms/try-phi3.5mini)
### Tokenizer
Phi-3.5-mini-Instruct supports a vocabulary size of up to `32064` tokens. The [tokenizer files](https://huggingface.co/microsoft/Phi-3.5-mini-instruct/blob/main/added_tokens.json) already provide placeholder tokens that can be used for downstream fine-tuning, but they can also be extended up to the model's vocabulary size.
### Input Formats
Given the nature of the training data, the Phi-3.5-mini-instruct model is best suited for prompts using the chat format as follows:
```
<|system|>
You are a helpful assistant.<|end|>
<|user|>
How to explain Internet for a medieval knight?<|end|>
<|assistant|>
```
### Loading the model locally
After obtaining the Phi-3.5-mini-instruct model checkpoint, users can use this sample code for inference.
```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
torch.random.manual_seed(0)
model = AutoModelForCausalLM.from_pretrained(
"microsoft/Phi-3.5-mini-instruct",
device_map="cuda",
torch_dtype="auto",
trust_remote_code=True,
)
tokenizer = AutoTokenizer.from_pretrained("microsoft/Phi-3.5-mini-instruct")
messages = [
{"role": "system", "content": "You are a helpful AI assistant."},
{"role": "user", "content": "Can you provide ways to eat combinations of bananas and dragonfruits?"},
{"role": "assistant", "content": "Sure! Here are some ways to eat bananas and dragonfruits together: 1. Banana and dragonfruit smoothie: Blend bananas and dragonfruits together with some milk and honey. 2. Banana and dragonfruit salad: Mix sliced bananas and dragonfruits together with some lemon juice and honey."},
{"role": "user", "content": "What about solving an 2x + 3 = 7 equation?"},
]
pipe = pipeline(
"text-generation",
model=model,
tokenizer=tokenizer,
)
generation_args = {
"max_new_tokens": 500,
"return_full_text": False,
"temperature": 0.0,
"do_sample": False,
}
output = pipe(messages, **generation_args)
print(output[0]['generated_text'])
```
Notes: If you want to use flash attention, call _AutoModelForCausalLM.from_pretrained()_ with _attn_implementation="flash_attention_2"_
## Responsible AI Considerations
Like other language models, the Phi family of models can potentially behave in ways that are unfair, unreliable, or offensive. Some of the limiting behaviors to be aware of include:
+ Quality of Service: The Phi models are trained primarily on English text and some additional multilingual text. Languages other than English will experience worse performance as well as performance disparities across non-English. English language varieties with less representation in the training data might experience worse performance than standard American English.
+ Multilingual performance and safety gaps: We believe it is important to make language models more widely available across different languages, but the Phi 3 models still exhibit challenges common across multilingual releases. As with any deployment of LLMs, developers will be better positioned to test for performance or safety gaps for their linguistic and cultural context and customize the model with additional fine-tuning and appropriate safeguards.
+ 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, cultural contexts, or prevalence of examples of negative stereotypes in training data that reflect real-world patterns and societal biases.
+ 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 case.
+ Information Reliability: Language models can generate nonsensical content or fabricate content that might sound reasonable but is inaccurate or outdated.
+ Limited Scope for Code: Majority of Phi-3 training data is based in Python and use 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.
+ Long Conversation: Phi-3 models, like other models, can in some cases generate responses that are repetitive, unhelpful, or inconsistent in very long chat sessions in both English and non-English languages. Developers are encouraged to place appropriate mitigations, like limiting conversation turns to account for the possible conversational drift
Developers should apply responsible AI best practices, including mapping, measuring, and mitigating risks associated with their specific use case and cultural, linguistic context. Phi-3 family of models are general purpose models. As developers plan to deploy these models for specific use cases, they are encouraged to fine-tune the models for their use case and leverage the models as part of broader AI systems with language-specific safeguards in place. Important areas for consideration include:
+ 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.
+ High-Risk Scenarios: Developers should assess the 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.
+ 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).
+ Generation of Harmful Content: Developers should assess outputs for their context and use available safety classifiers or custom solutions appropriate for their use case.
+ 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.
## Training
### Model
**Architecture:** Phi-3.5-mini has 3.8B parameters and is a dense decoder-only Transformer model using the same tokenizer as Phi-3 Mini.<br>
**Inputs:** Text. It is best suited for prompts using chat format.<br>
**Context length:** 128K tokens<br>
**GPUs:** 512 H100-80G<br>
**Training time:** 10 days<br>
**Training data:** 3.4T tokens<br>
**Outputs:** Generated text in response to the input<br>
**Dates:** Trained between June and August 2024<br>
**Status:** This is a static model trained on an offline dataset with cutoff date October 2023 for publicly available data. Future versions of the tuned models may be released as we improve models.<br>
**Supported languages:** Arabic, Chinese, Czech, Danish, Dutch, English, Finnish, French, German, Hebrew, Hungarian, Italian, Japanese, Korean, Norwegian, Polish, Portuguese, Russian, Spanish, Swedish, Thai, Turkish, Ukrainian<br>
**Release date:** August 2024<br>
# Uploaded model
- **Developed by:** EpistemeAI
- **License:** apache-2.0
- **Finetuned from model :** unsloth/phi-3.5-mini-instruct-bnb-4bit
This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth) |