Text Generation
Transformers
Safetensors
flint
Mixture of Experts
gptq
int8
w8a8
gs128
ptqr
quantized
vllm
amd
mi100
gfx908
rocm
conversational
8-bit precision
Instructions to use voodooquant/Flint-GPTQ-INT8-PTQR with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use voodooquant/Flint-GPTQ-INT8-PTQR with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="voodooquant/Flint-GPTQ-INT8-PTQR") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("voodooquant/Flint-GPTQ-INT8-PTQR") model = AutoModelForCausalLM.from_pretrained("voodooquant/Flint-GPTQ-INT8-PTQR", 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=256) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use voodooquant/Flint-GPTQ-INT8-PTQR with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "voodooquant/Flint-GPTQ-INT8-PTQR" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "voodooquant/Flint-GPTQ-INT8-PTQR", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/voodooquant/Flint-GPTQ-INT8-PTQR
- SGLang
How to use voodooquant/Flint-GPTQ-INT8-PTQR 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 "voodooquant/Flint-GPTQ-INT8-PTQR" \ --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": "voodooquant/Flint-GPTQ-INT8-PTQR", "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 "voodooquant/Flint-GPTQ-INT8-PTQR" \ --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": "voodooquant/Flint-GPTQ-INT8-PTQR", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use voodooquant/Flint-GPTQ-INT8-PTQR with Docker Model Runner:
docker model run hf.co/voodooquant/Flint-GPTQ-INT8-PTQR
Download quantize_config.json from voodooquant/Flint-GPTQ-INT8-PTQR: direct link, hf CLI and curl.
- Browser
- Download file 286 Bytes
-
https://huggingface.co/voodooquant/Flint-GPTQ-INT8-PTQR/resolve/main/quantize_config.json
- Command line
-
hf download hf://voodooquant/Flint-GPTQ-INT8-PTQR/quantize_config.json
-
curl -L -o quantize_config.json https://huggingface.co/voodooquant/Flint-GPTQ-INT8-PTQR/resolve/main/quantize_config.json
286 Bytes
| { | |
| "bits": 8, | |
| "group_size": 128, | |
| "desc_act": false, | |
| "sym": true, | |
| "lm_head": false, | |
| "method": "gptq", | |
| "quant_method": "gptq", | |
| "checkpoint_format": "gptq", | |
| "pack_dtype": "int32", | |
| "meta": { | |
| "quantizer": [ | |
| "ptqr-retrain:scripts/ptqr_export_flint.py" | |
| ] | |
| } | |
| } |