Instructions to use voidful/llmcodec-librispeech with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use voidful/llmcodec-librispeech with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("voidful/llm-codec")
model = PeftModel.from_pretrained(base_model, "voidful/llmcodec-librispeech")

Transformers

How to use voidful/llmcodec-librispeech with Transformers:

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

pipe = pipeline("text-generation", model="voidful/llmcodec-librispeech")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("voidful/llmcodec-librispeech")
model = AutoModelForCausalLM.from_pretrained("voidful/llmcodec-librispeech")
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

vLLM

How to use voidful/llmcodec-librispeech with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/voidful/llmcodec-librispeech

SGLang

How to use voidful/llmcodec-librispeech 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 "voidful/llmcodec-librispeech" \
    --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": "voidful/llmcodec-librispeech",
		"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 "voidful/llmcodec-librispeech" \
        --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": "voidful/llmcodec-librispeech",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use voidful/llmcodec-librispeech with Docker Model Runner:
```
docker model run hf.co/voidful/llmcodec-librispeech
```

llmcodec-librispeech / README.md

voidful

Model save

7b8384a verified 5 months ago

preview code

raw

history blame contribute delete

1.56 kB

metadata

library_name: peft
base_model: voidful/llm-codec
tags:
  - base_model:adapter:voidful/llm-codec
  - lora
  - transformers
pipeline_tag: text-generation
model-index:
  - name: llmcodec-librispeech
    results: []

llmcodec-librispeech

This model is a fine-tuned version of voidful/llm-codec on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 8.7129

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0001
train_batch_size: 1
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 8
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.1
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
8.8157	1.0	35156	8.7129

Framework versions

PEFT 0.18.0
Transformers 4.57.3
Pytorch 2.9.0.dev20250629+cu128
Datasets 2.14.6
Tokenizers 0.22.1