Instructions to use ybelkada/bloom-1b7-8bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use ybelkada/bloom-1b7-8bit with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="ybelkada/bloom-1b7-8bit")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("ybelkada/bloom-1b7-8bit") model = AutoModelForCausalLM.from_pretrained("ybelkada/bloom-1b7-8bit") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use ybelkada/bloom-1b7-8bit with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "ybelkada/bloom-1b7-8bit" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "ybelkada/bloom-1b7-8bit", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/ybelkada/bloom-1b7-8bit
- SGLang
How to use ybelkada/bloom-1b7-8bit 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 "ybelkada/bloom-1b7-8bit" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "ybelkada/bloom-1b7-8bit", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'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 "ybelkada/bloom-1b7-8bit" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "ybelkada/bloom-1b7-8bit", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use ybelkada/bloom-1b7-8bit with Docker Model Runner:
docker model run hf.co/ybelkada/bloom-1b7-8bit
Upload BloomForCausalLM
Browse files- config.json +8 -0
config.json
CHANGED
|
@@ -21,6 +21,14 @@
|
|
| 21 |
"offset_alibi": 100,
|
| 22 |
"pad_token_id": 3,
|
| 23 |
"pretraining_tp": 2,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
"seq_length": 4096,
|
| 25 |
"skip_bias_add": true,
|
| 26 |
"skip_bias_add_qkv": false,
|
|
|
|
| 21 |
"offset_alibi": 100,
|
| 22 |
"pad_token_id": 3,
|
| 23 |
"pretraining_tp": 2,
|
| 24 |
+
"quantization_config": {
|
| 25 |
+
"_from_model_config": false,
|
| 26 |
+
"llm_int8_enable_fp32_cpu_offload": false,
|
| 27 |
+
"llm_int8_skip_modules": null,
|
| 28 |
+
"llm_int8_threshold": 6.0,
|
| 29 |
+
"load_in_8bit": true,
|
| 30 |
+
"transformers_version": "4.28.0.dev0"
|
| 31 |
+
},
|
| 32 |
"seq_length": 4096,
|
| 33 |
"skip_bias_add": true,
|
| 34 |
"skip_bias_add_qkv": false,
|