GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
EMTIAZZ
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
EMTIAZZ
Model Tree
canopylabs/orpheus-3b-0.1-pretrained
Input Modalities
Output Modalities
Supported Functionality
| Property | Value |
|---|---|
| Base Model | canopylabs/orpheus-3b-0.1-pretrained |
| Architecture | Llama 3B + LoRA adapters |
| Training Data | ~99,000 Bangla speech samples (combined dataset) |
| Training Steps | 30,000 |
| Audio Codec | SNAC 24kHz |
| Training Platform | Modal (H100 GPU) with Unsloth |
| Language | Bangla (bn) |
| License | Apache 2.0 |
Orpheus TTS is a Llama-based text-to-speech model that generates audio as interleaved SNAC codec tokens. It supports emotional speech tags like <laugh>, <sigh>, <gasp>, etc.
Note: The base model
canopylabs/orpheus-3b-0.1-pretrainedis gated — you need a HuggingFace token with access approved.
python
import torchfrom transformers import AutoTokenizer, AutoModelForCausalLMfrom peft import PeftModelimport snacimport numpy as npimport soundfile as sf # 1. Load base modelbase_model_id = "canopylabs/orpheus-3b-0.1-pretrained"tokenizer = AutoTokenizer.from_pretrained(base_model_id, token="YOUR_HF_TOKEN")model = AutoModelForCausalLM.from_pretrained( base_model_id, torch_dtype=torch.float16, token="YOUR_HF_TOKEN") # 2. Resize embeddings (vocab size mismatch fix)model.resize_token_embeddings(156940) # 3. Load LoRA adaptermodel = PeftModel.from_pretrained( model, "EMTIAZZ/orpheus-3b-bangla-high-data-finetuning", token="YOUR_HF_TOKEN")model = model.merge_and_unload() # 4. Generate speechtext = "আমি বাংলায় কথা বলতে পারি।"prompt = f"<|begin_of_text|><|start_header_id|>user<|end_header_id|>\n\n{text}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n" inputs = tokenizer(prompt, return_tensors="pt")with torch.no_grad(): outputs = model.generate( **inputs, max_new_tokens=1200, do_sample=True, temperature=0.6, top_p=0.95, repetition_penalty=1.1, )The model supports speech emotion/style tags (inherited from Orpheus):
text
<laugh> <chuckle> <sigh> <cough> <sniffle><groan> <yawn> <gasp>Example: "সে বলল <laugh> এটা সত্যিই মজার!"
| Model | Training Data | Steps | Quality |
|---|---|---|---|
| This model (high data) | ~99K samples | 30,000 | Better prosody, more natural |
| Small data version | ~39K samples | 4,500 | Faster to infer, lighter |
If you use this model, please cite:
bibtex
@misc{emtiaz2026orpheusbangla, author = {Emtiaz Uddin Ahmed}, title = {Orpheus 3B Bangla High-Data Fine-tune}, year = {2026}, url = {https://huggingface.co/EMTIAZZ/orpheus-3b-bangla-high-data-finetuning}}import torchfrom transformers import AutoTokenizer, AutoModelForCausalLMfrom peft import PeftModelimport snacimport numpy as npimport soundfile as sf # 1. Load base modelbase_model_id = "canopylabs/orpheus-3b-0.1-pretrained"tokenizer = AutoTokenizer.from_pretrained(base_model_id, token="YOUR_HF_TOKEN")model = AutoModelForCausalLM.from_pretrained( base_model_id, torch_dtype=torch.float16, token="YOUR_HF_TOKEN") # 2. Resize embeddings (vocab size mismatch fix)model.resize_token_embeddings(156940) # 3. Load LoRA adaptermodel = PeftModel.from_pretrained( model, "EMTIAZZ/orpheus-3b-bangla-high-data-finetuning", token="YOUR_HF_TOKEN")model = model.merge_and_unload() # 4. Generate speechtext = "আমি বাংলায় কথা বলতে পারি।"prompt = f"<|begin_of_text|><|start_header_id|>user<|end_header_id|>\n\n{text}<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n" inputs = tokenizer(prompt, return_tensors="pt")with torch.no_grad(): outputs = model.generate( **inputs, max_new_tokens=1200, do_sample=True, temperature=0.6, top_p=0.95, repetition_penalty=1.1, )@misc{emtiaz2026orpheusbangla, author = {Emtiaz Uddin Ahmed}, title = {Orpheus 3B Bangla High-Data Fine-tune}, year = {2026}, url = {https://huggingface.co/EMTIAZZ/orpheus-3b-bangla-high-data-finetuning}}