Update README.md

README.md

@@ -1,199 +1,113 @@
 ---
+language:
+- en
+- ar
+- de
+- es
+- fr
+- it
+- ja
+- pt
+- ru
+license: mit
 library_name: transformers
-tags: []
+datasets:
+- fixie-ai/librispeech_asr
+- fixie-ai/common_voice_17_0
+- fixie-ai/peoples_speech
+- fnlp/AnyInstruct
+metrics:
+- bleu
 ---
 
-# Model Card for Model ID
+# Model Card for Ultravox
 
-<!-- Provide a quick summary of what the model is/does. -->
+Ultravox is a multimodal Speech LLM built around a pretrained [Llama3.1-70B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B) and [Whisper-medium](https://huggingface.co/openai/whisper-medium) backbone.
 
+See https://ultravox.ai for the GitHub repo and more information.
 
 
 ## Model Details
 
 ### Model Description
 
-<!-- Provide a longer summary of what this model is. -->
+Ultravox is a multimodal model that can consume both speech and text as input (e.g., a text system prompt and voice user message). 
+The input to the model is given as a text prompt with a special `<|audio|>` pseudo-token, and the model processor will replace this magic token with embeddings derived from the input audio.
+Using the merged embeddings as input, the model will then generate output text as usual. 
 
-This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
+In a future revision of Ultravox, we plan to expand the token vocabulary to support generation of semantic and acoustic audio tokens, which can then be fed to a vocoder to produce voice output.
+No preference tuning has been applied to this revision of the model.
 
-- **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]
+- **Developed by:** Fixie.ai
+- **License:** MIT
 
-### Model Sources [optional]
+### Model Sources
 
-<!-- Provide the basic links for the model. -->
+- **Repository:** https://ultravox.ai
+- **Demo:** See repo
 
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
+## Usage
 
-## Uses
+Think of the model as an LLM that can also hear and understand speech. As such, it can be used as a voice agent, and also to do speech-to-speech translation, analysis of spoken audio, etc.
 
-<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+To use the model, try the following:
+```python
+# pip install transformers peft librosa
 
-### Direct Use
+import transformers
+import numpy as np
+import librosa
 
-<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+pipe = transformers.pipeline(model='fixie-ai/ultravox-v0_4', trust_remote_code=True)
 
-[More Information Needed]
+path = "<path-to-input-audio>"  # TODO: pass the audio here
+audio, sr = librosa.load(path, sr=16000)
 
-### Downstream Use [optional]
 
-<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+turns = [
+  {
+    "role": "system",
+    "content": "You are a friendly and helpful character. You love to answer questions for people."
+  },
+]
+pipe({'audio': audio, 'turns': turns, 'sampling_rate': sr}, max_new_tokens=30)
+```
 
-[More Information Needed]
 
-### Out-of-Scope Use
-
-<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
-
-[More Information Needed]
-
-## Bias, Risks, and Limitations
-
-<!-- This section is meant to convey both technical and sociotechnical limitations. -->
-
-[More Information Needed]
-
-### Recommendations
-
-<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
-
-Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
-
-## How to Get Started with the Model
+## Training Details
 
-Use the code below to get started with the model.
+The model uses a pre-trained [Llama3.1-70B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-70B) backbone as well as the encoder part of [Whisper-medium](https://huggingface.co/openai/whisper-medium).
 
-[More Information Needed]
+Only the multi-modal adapter is trained, while Whisper encoder and Llama are kept frozen.
 
-## Training Details
+We use a knowledge-distillation loss where Ultravox is trying to match the logits of the text-based Llama backbone.
 
 ### Training Data
 
-<!-- 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. -->
+Training dataset is a mix of ASR datasets, extended by adding a "continuation" generated by Llama 3.1 70B.
 
-[More Information Needed]
 
 ### Training Procedure
 
-<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
-
-#### Preprocessing [optional]
-
-[More Information Needed]
+Supervised speech to audio finetuning. For more info, see [training code in Ultravox repo](https://github.com/fixie-ai/ultravox/blob/main/ultravox/training/train.py).
 
 
 #### Training Hyperparameters
 
-- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+- **Training regime:** BF16 mixed precision training
+- **Hardward used:** 8x H100 GPUs
 
-#### Speeds, Sizes, Times [optional]
+#### Speeds, Sizes, Times
 
-<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+The current version of Ultravox, when invoked with audio content, has a time-to-first-token (TTFT) of approximately 400ms, and a tokens-per-second rate of ~50-100 when using 4xA100-40GB GPU, all using a Llama 3.1 70B backbone.
 
-[More Information Needed]
+Check out the audio tab on [TheFastest.ai](https://thefastest.ai/?m=audio) for daily benchmarks and a comparison with other existing models.
 
 ## Evaluation
-
-<!-- This section describes the evaluation protocols and provides the results. -->
-
-### Testing Data, Factors & Metrics
-
-#### Testing Data
-
-<!-- This should link to a Dataset Card if possible. -->
-
-[More Information Needed]
-
-#### Factors
-
-<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
-
-[More Information Needed]
-
-#### Metrics
-
-<!-- These are the evaluation metrics being used, ideally with a description of why. -->
-
-[More Information Needed]
-
-### Results
-
-[More Information Needed]
-
-#### Summary
-
-
-
-## 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]
+|                   | en_de (BLEU) | es_en (BLEU) | LibriSpeech clean.test (WER) |
+|:------------------|:-------------|:-------------|:----------------------------|
+| Ultravox v0.3 | 22.66        | 24.74        | 6.67                        |
+| Ultravox v0.4 8B | 25.47        | 37.11        | 4.45                        |
+| **Ultravox v0.4 70B** | **30.30**        | **39.55**        | **4.49**                        |
+| Llama3.1 8B (text-only) | 31.95     | 44.62        | -                           |
+| Llama3.1 70B (text-only) | 32.59     | 46.39        | -                           |