Instructions to use bigscience/distill-bloom-1b3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use bigscience/distill-bloom-1b3 with Transformers:

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

pipe = pipeline("text-generation", model="bigscience/distill-bloom-1b3")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("bigscience/distill-bloom-1b3")
model = AutoModelForCausalLM.from_pretrained("bigscience/distill-bloom-1b3")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use bigscience/distill-bloom-1b3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "bigscience/distill-bloom-1b3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "bigscience/distill-bloom-1b3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/bigscience/distill-bloom-1b3

SGLang

How to use bigscience/distill-bloom-1b3 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 "bigscience/distill-bloom-1b3" \
    --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": "bigscience/distill-bloom-1b3",
		"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 "bigscience/distill-bloom-1b3" \
        --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": "bigscience/distill-bloom-1b3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use bigscience/distill-bloom-1b3 with Docker Model Runner:
```
docker model run hf.co/bigscience/distill-bloom-1b3
```

Fix architecture

by lewtun HF Staff - opened Feb 10, 2023

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

-2

Files changed (1) hide show

config.json +2 -2

config.json CHANGED Viewed

@@ -1,7 +1,7 @@
 {
   "apply_residual_connection_post_layernorm": false,
   "architectures": [
-    "BloomModel"
   ],
   "attention_dropout": 0.0,
   "attention_softmax_in_fp32": true,
@@ -32,4 +32,4 @@
   "vocab_size": 250880,
   "weights_aggregation_strategy": "mean",
   "width_downsampling_rate": 0.5
-}

 {
   "apply_residual_connection_post_layernorm": false,
   "architectures": [
+    "BloomForCausalLM"
   ],
   "attention_dropout": 0.0,
   "attention_softmax_in_fp32": true,
   "vocab_size": 250880,
   "weights_aggregation_strategy": "mean",
   "width_downsampling_rate": 0.5
+}