Instructions to use hypaai/Hypa-Orpheus-3b-TTS-VC-ext with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use hypaai/Hypa-Orpheus-3b-TTS-VC-ext with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="hypaai/Hypa-Orpheus-3b-TTS-VC-ext")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("hypaai/Hypa-Orpheus-3b-TTS-VC-ext")
model = AutoModelForCausalLM.from_pretrained("hypaai/Hypa-Orpheus-3b-TTS-VC-ext", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use hypaai/Hypa-Orpheus-3b-TTS-VC-ext with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "hypaai/Hypa-Orpheus-3b-TTS-VC-ext"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "hypaai/Hypa-Orpheus-3b-TTS-VC-ext",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/hypaai/Hypa-Orpheus-3b-TTS-VC-ext

SGLang

How to use hypaai/Hypa-Orpheus-3b-TTS-VC-ext with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "hypaai/Hypa-Orpheus-3b-TTS-VC-ext" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "hypaai/Hypa-Orpheus-3b-TTS-VC-ext",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "hypaai/Hypa-Orpheus-3b-TTS-VC-ext" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "hypaai/Hypa-Orpheus-3b-TTS-VC-ext",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio

How to use hypaai/Hypa-Orpheus-3b-TTS-VC-ext with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for hypaai/Hypa-Orpheus-3b-TTS-VC-ext to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for hypaai/Hypa-Orpheus-3b-TTS-VC-ext to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for hypaai/Hypa-Orpheus-3b-TTS-VC-ext to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="hypaai/Hypa-Orpheus-3b-TTS-VC-ext",
    max_seq_length=2048,
)

Docker Model Runner
How to use hypaai/Hypa-Orpheus-3b-TTS-VC-ext with Docker Model Runner:
```
docker model run hf.co/hypaai/Hypa-Orpheus-3b-TTS-VC-ext
```

Extended continued fine-tune of Hypa-Orpheus 3B with a higher-capacity LoRA adapter (r=2048), merged to 16-bit bf16 for inference.

Model Description

Hypa-Orpheus 3B Extended (hypaai/Hypa-Orpheus-3b-TTS-VC-ext) is a continued fine-tune of the production Hypa-Orpheus 3B release. It uses the same Hypa-Voices corpus, prompt formats, and SNAC training pipeline as the primary model, but attaches a much higher-rank LoRA adapter (r=2048, α=2048) on top of the merged Step 3a checkpoint and trains further on the same precomputed integer SNAC shards.

This is an experimental capacity extension, not a replacement for the primary release. It exists for researchers and deployers who want the strongest held-out loss we observed after continuing training with a larger adapter.

Collection: hypaai/hypa-orpheus

Repository	Role	Best for
`hypaai/Hypa-Orpheus-3b-TTS-VC`	Primary merged release (r=512, step 44,334)	Default production deployment
`hypaai/Hypa-Orpheus-3b-TTS-VC-ext` (this repo)	Extended merged release (r=2048, step 4,992)	Higher-capacity continuation checkpoint
`hypaai/Hypa-Orpheus-3b-TTS-VC-LoRAs`	Primary LoRA + TensorBoard	Primary adapter inspection and retraining
`hypaai/Hypa-Orpheus-3b-TTS-VC-LoRAs-ext`	Extended LoRA + TensorBoard	Extended adapter inspection and retraining

Property	Primary release	Extended (this repo)
Starting checkpoint	`unsloth/orpheus-3b-0.1-ft`	`hypaai/Hypa-Orpheus-3b-TTS-VC`
LoRA rank / alpha	r=512, α=512	r=2048, α=2048
Trainable parameters	778M (19.07%)	~~3.11B (~~76%)
Adapter size	~3.1 GB	~12.4 GB
Effective batch size	96	256
Optimizer steps (shipped)	44,334	4,992
Best validation loss	3.513	3.438
Approx. validation PPL	33.5	31.1

Capabilities, languages, prompt structure, SNAC decode path, and intended use cases are identical to the primary model. See the main model card for the full description:

hypaai/Hypa-Orpheus-3b-TTS-VC README

What Changed in This Run

The extended run kept the same data mixture and response-only SNAC supervision, but increased adapter capacity and batch size while continuing from the already merged production weights:

Hyperparameter	Primary release	Extended run
Base weights	4-bit QLoRA on Orpheus FT	Merged bf16 Hypa-Orpheus 3B
LoRA rank (r)	512	2048
LoRA alpha (α)	512	2048
Per-device batch size	16	16
Gradient accumulation	6	16
Effective batch size	96	256
Learning rate	5e-5	5e-5
LR schedule	cosine, 3% warmup	cosine, 498 warmup steps
Optimizer	AdamW 8-bit	AdamW 8-bit
Weight decay	0.01	0.01
Max grad norm	1.0	1.0
Eval / save interval	2,217 steps	832 steps
Planned step budget	44,334	16,626
Shipped checkpoint	step 44,334	step 4,992
Precision	bfloat16	bfloat16
Random seed	3407	3407

The published extended checkpoint reflects 4,992 optimizer steps, about 30% of one effective epoch on the same interleaved TTS/VC mixture. Training was stopped and released at the validation-best step observed in that continuation segment.

Evaluation and Checkpoint Selection

TensorBoard logs for the extended adapter:

Hypa-Orpheus 3B Extended LoRA training logs

Headline metrics from the shipped continuation:

Metric	Value	Step
Initial training loss	3.582	1
Final training loss	3.433	4,992
Initial validation loss	3.496	832
Best validation loss (shipped)	3.438	4,992

Compared with the primary release validation loss of 3.513 at step 44,334, the extended checkpoint improves held-out loss by 0.075 (approximate validation perplexity 33.5 → 31.1). Validation loss improved at every logged evaluation interval in this continuation.

Step	Training Loss	Validation Loss	Val PPL	Gen. Gap
832	3.531	3.496	33.0	−0.035
1,664	3.491	3.478	32.4	−0.013
2,496	3.483	3.464	31.9	−0.019
3,328	3.469	3.456	31.7	−0.012
4,160	3.463	3.451	31.5	−0.012
4,992	3.433	3.438	31.1	+0.006

As with the primary release, these scalars measure uncertainty over the supervised target SNAC stream. They do not, by themselves, prove better perceptual quality. The extended checkpoint should be treated as a loss-improved continuation pending formal listening and task benchmarks.

Gradient norms remained stable throughout the run (roughly 0.18 → 0.11).

Which Checkpoint Should I Use?

Use case	Recommendation
Default production	`hypaai/Hypa-Orpheus-3b-TTS-VC`
Smaller adapter / full primary training budget	Primary LoRA repo
Lowest published held-out loss after continuation	This repo (`-ext`)
Further high-rank experimentation	Extended LoRA repo

Choose the primary release when you want the fully budgeted Step 3a training run and the smaller r=512 adapter footprint. Choose this extended release when you want the best validation loss we published after continuing from that checkpoint with a r=2048 adapter, and can accept the larger merged weights and shorter continuation schedule.

Quick Start

Load the extended merged model

from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

model_id = "hypaai/Hypa-Orpheus-3b-TTS-VC-ext"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.bfloat16,
    device_map="auto",
)

Prompt assembly, SNAC encode/decode, and vLLM serving follow the same conventions as the primary release. This repository includes the same handler.py inference entrypoint pattern as hypaai/Hypa-Orpheus-3b-TTS-VC.

vllm serve hypaai/Hypa-Orpheus-3b-TTS-VC-ext --dtype bfloat16 --max-model-len 4096

Load the extended LoRA adapter

See hypaai/Hypa-Orpheus-3b-TTS-VC-LoRAs-ext for adapter loading, merging, and TensorBoard links.

Limitations (Extended Release)

Shorter continuation budget. The shipped extended checkpoint covers 4,992 steps, not the full planned 16,626-step schedule.
Much larger adapter. r=2048 increases adapter storage and merge size roughly 4× versus the primary release.
Experimental status. The primary release remains our default production artifact; this continuation is published for transparency and research comparison.
Same perceptual eval gap. Improved held-out loss does not yet come with published WER/CER, speaker-similarity, or MOS results.

For shared limitations on languages, prompt sensitivity, and context length, see the primary model card.

Released Artifacts

Extended merged 16-bit model (this repo): hypaai/Hypa-Orpheus-3b-TTS-VC-ext
Extended LoRA + TensorBoard: hypaai/Hypa-Orpheus-3b-TTS-VC-LoRAs-ext
Primary merged release: hypaai/Hypa-Orpheus-3b-TTS-VC
Primary LoRA + TensorBoard: hypaai/Hypa-Orpheus-3b-TTS-VC-LoRAs
Training data (public subset): Hypa-Voices collection

For questions or contributions, contact chris@hypaintelligence.com or open an issue on Hugging Face.

Citation

If you use the extended Hypa-Orpheus checkpoint, please cite both the primary release and this continuation:

@misc{hypaorpheus2026,
  title        = {Hypa-Orpheus 3B: Multilingual Text-to-Speech and Voice Cloning for Underrepresented Languages},
  author       = {{Hypa Intelligence and AfroVoices}},
  year         = {2026},
  publisher    = {Hugging Face},
  howpublished = {\url{https://huggingface.co/hypaai/Hypa-Orpheus-3b-TTS-VC}}
}

@misc{hypaorpheus2026ext,
  title        = {Hypa-Orpheus 3B Extended: Continued Fine-Tuning with Higher-Capacity LoRA},
  author       = {{Hypa Intelligence and AfroVoices}},
  year         = {2026},
  publisher    = {Hugging Face},
  howpublished = {\url{https://huggingface.co/hypaai/Hypa-Orpheus-3b-TTS-VC-ext}}
}

License

Released under the Apache License 2.0.

Acknowledgments

Same upstream acknowledgments as the primary Hypa-Orpheus release: Canopy Labs (Orpheus), Meta AI (Llama), Hubert Siuzdak (SNAC), Unsloth, RunPod, Mozilla Common Voice, AfroVoices, and Hypa Intelligence Research (HaIR).

Hypa Intelligence • Website • Hugging Face • Updates