mesolitica
/

Malaysian-Qwen2.5-7B-Speech-Instruct

@@ -1,199 +1,133 @@
 ---
-library_name: transformers
-tags: []
 ---
-# Model Card for Model ID
-<!-- Provide a quick summary of what the model is/does. -->
-## Model Details
-### Model Description
-<!-- Provide a longer summary of what this model is. -->
-This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
-- **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 Sources [optional]
-<!-- Provide the basic links for the model. -->
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
-## Uses
-<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
-### Direct Use
-<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
-[More Information Needed]
-### Downstream Use [optional]
-<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
-[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
-Use the code below to get started with the model.
-[More Information Needed]
-## Training Details
-### 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. -->
-[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]
-#### 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 -->
-#### Speeds, Sizes, Times [optional]
-<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
-[More Information Needed]
-## 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]

 ---
+language:
+- ms
+- en
+- zh
+base_model:
+- mesolitica/Malaysian-Qwen2.5-7B-Audio-Instruct
 ---
+# Malaysian-Audio-Qwen2.5-7B-Speech-Instruct
+Speech model on top of [mesolitica/Malaysian-Qwen2.5-7B-Audio-Instruct](https://huggingface.co/mesolitica/Malaysian-Qwen2.5-7B-Audio-Instruct). It designs for various audio tasks and suitable for voice assistant general question answer.
+## How we trained it?
+Speech instructions, actual conversations related to coding, politics, chat assistant and general QA.
+- We use freezed Whisper Large V3 Encoder without any pooling, means 30 seconds audio consumed 1500 tokens or 1 token equal to 0.02 seconds.
+- Projection, Embedding and LM Head layers are done in full parameter finetuning.
+- LoRA for other linear layers with rank 64 and alpha 128.
+- Training done in multipacking with 10240 context length.
+- WanDB at https://wandb.ai/huseinzol05/lora-embedding-64-audio-qwen2.5-7b-malaysian-10k-stage2
+- Revision [513a900f40d372e8d7eb774e0561af043c704449](https://huggingface.co/mesolitica/Malaysian-Audio-Qwen2.5-7B-Instruct/commit/513a900f40d372e8d7eb774e0561af043c704449)
+#### Dataset
+1. [mesolitica/Malaysian-UltraChat-Speech-Multiturn-Instructions](https://huggingface.co/datasets/mesolitica/Malaysian-UltraChat-Speech-Multiturn-Instructions), 1 epoch.
+2. [mesolitica/Malaysian-Multiturn-Chat-Assistant](https://huggingface.co/datasets/mesolitica/Malaysian-Multiturn-Chat-Assistant), 1 epoch.
+3. [mesolitica/Malaysian-Speech-Instructions](https://huggingface.co/datasets/mesolitica/Malaysian-Speech-Instructions), 1 epoch.
+4. [mesolitica/Malaysian-Reasoning-Speech-Instructions](https://huggingface.co/datasets/mesolitica/Malaysian-Reasoning-Speech-Instructions), 1 epoch.
+5. [mesolitica/Malaysian-Speech-Description-Timestamp-Instructions](https://huggingface.co/datasets/mesolitica/Malaysian-Speech-Description-Timestamp-Instructions), random sampling 0.2 epoch.
+6. [mesolitica/Cantonese-Radio-Description-Instructions](https://huggingface.co/datasets/mesolitica/Cantonese-Radio-Description-Instructions), random sampling 0.2 epoch.
+7. [mesolitica/Emilia-Mandarin-Description-Instructions](https://huggingface.co/datasets/mesolitica/Emilia-Mandarin-Description-Instructions), random sampling 0.2 epoch.
+8. [mesolitica/Malaysian-SFT/combined-malaysian-sft-5k-sample.jsonl](https://huggingface.co/datasets/mesolitica/Malaysian-SFT/blob/main/combine/combined-malaysian-sft-5k-sample.jsonl), text corpus, 1 epoch.
+9. [mesolitica/Malaysian-Instructions/voice_assistant](https://huggingface.co/datasets/mesolitica/Malaysian-Instructions/viewer/default/voice_assistant), text only instructions, 1 epoch.
+10. [mesolitica/Malaysian-Instructions/mixed_manglish](https://huggingface.co/datasets/mesolitica/Malaysian-Instructions/viewer/default/mixed_manglish), text only instructions, 1 epoch.
+11. [mesolitica/Malaysian-Instructions/manglish](https://huggingface.co/datasets/mesolitica/Malaysian-Instructions/viewer/default/manglish), text only instructions, 1 epoch.
+12. [mesolitica/Malaysian-Instructions/longer_respond](https://huggingface.co/datasets/mesolitica/Malaysian-Instructions/viewer/default/longer_respond), text only instructions, 1 epoch.
+With total 3.14B tokens (include text only instructions) or 9584.595 audio hours.
+## Benchmark
+## How to use
+```python
+from transformers import AutoFeatureExtractor, AutoModelForCausalLM, AutoTokenizer
+from transformers import TextStreamer
+import torch
+import librosa
+import math
+def process(messages, device = 'cpu', dtype = torch.float32):
+    audios = []
+    for message in messages:
+        if isinstance(message["content"], list):
+            for ele in message["content"]:
+                if ele["type"] == "audio":
+                    audios.append(ele['audio_url'])
+    y = []
+    audio_length = 0
+    for f in audios:
+        y_, _ = librosa.load(f, sr = 16000)
+        y.append(y_)
+        audio_length += min(3000, math.ceil(len(y_) / feature_extractor.hop_length))
+    audio_length = (audio_length - 1) // 2 + 1
+    text = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
+    expanded_audio_token = audio_token * audio_length
+    text = text.replace(audio_token, expanded_audio_token)
+    inputs = tokenizer(text, return_tensors = 'pt')
+    input_ids = inputs['input_ids']
+    inputs_audio = feature_extractor(
+        y,
+        return_attention_mask=True,
+        padding="max_length",
+        sampling_rate=16000,
+        return_tensors = 'pt'
+    )
+    input_features = inputs_audio['input_features']
+    feature_attention_mask = inputs_audio['attention_mask']
+    return {
+        'input_ids': input_ids.to(device),
+        'input_features': input_features.to(dtype).to(device),
+        'feature_attention_mask': feature_attention_mask.to(device)
+    }
+tokenizer = AutoTokenizer.from_pretrained('mesolitica/Malaysian-Qwen2.5-7B-Speech-Instruct')
+feature_extractor = AutoFeatureExtractor.from_pretrained('mesolitica/Malaysian-Qwen2.5-7B-Speech-Instruct')
+streamer = TextStreamer(tokenizer)
+audio_token = "<|AUDIO|>"
+model = AutoModelForCausalLM.from_pretrained(
+    'mesolitica/Malaysian-Qwen2.5-7B-Speech-Instruct',
+    torch_dtype = 'auto', trust_remote_code = True,
+).cuda()
+messages = [
+    {"role": "user", "content": [
+        {"type": "audio", "audio_url": 'speech/coding-pca-pytorch.mp3'},
+        {"type": "text", "text": "is the audio about chicken"},
+    ]},
+]
+inputs = process(messages, device = model.device, dtype = model.dtype)
+with torch.no_grad():
+    generate_kwargs = dict(
+        **inputs,
+        max_new_tokens=512,
+        top_p=0.95,
+        top_k=50,
+        temperature=0.9,
+        do_sample=True,
+        repetition_penalty=1.05,
+        streamer=streamer
+    )
+    generation_output = model.generate(**generate_kwargs)
+```
+Output,
+```
+<|im_start|>system
+You are a helpful assistant.<|im_end|>
+<|im_start|>user
+Audio 1: <|audio_bos|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|AUDIO|><|audio_eos|>
+is the audio about chicken<|im_end|>
+<|im_start|>assistant
+The audio is a conversation where someone asks about applying PyTorch for implementing dimensionality reduction techniques with Principal Component Analysis (PCA). The context suggests an academic or technical discussion, possibly in a classroom or seminar setting. The tone of the speaker is likely educational and engaged, seeking to understand how to effectively use PyTorch for data analysis tasks involving dimensionality reduction.<|im_end|>
+```
+**The generation is stochastic, expect some randomness when run locally**.
+You can try more audio examples at https://github.com/mesolitica/malaya/tree/master/session/audiollm/speech