Instructions to use power-greg/super-fast-llm with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use power-greg/super-fast-llm with Transformers:

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

pipe = pipeline("text-generation", model="power-greg/super-fast-llm")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("power-greg/super-fast-llm")
model = AutoModelForCausalLM.from_pretrained("power-greg/super-fast-llm")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use power-greg/super-fast-llm with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/power-greg/super-fast-llm

SGLang

How to use power-greg/super-fast-llm 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 "power-greg/super-fast-llm" \
    --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": "power-greg/super-fast-llm",
		"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 "power-greg/super-fast-llm" \
        --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": "power-greg/super-fast-llm",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use power-greg/super-fast-llm with Docker Model Runner:
```
docker model run hf.co/power-greg/super-fast-llm
```

power-greg commited on Dec 1, 2022

Commit

a8e76d5

1 Parent(s): bff5280

Upload 2 files

Browse files

Files changed (2) hide show

config.json +5 -5
pytorch_model.bin +2 -2

config.json CHANGED Viewed

@@ -1,10 +1,10 @@
 {
-  "_name_or_path": "/tmp/tmpcaxm4ggs",
   "activation_function": "gelu_new",
   "architectures": [
     "GPT2LMHeadModel"
   ],
-  "attn_pdrop": 0.2,
   "bos_token_id": 0,
   "embd_pdrop": 0.0,
   "eos_token_id": 0,
@@ -12,9 +12,9 @@
   "layer_norm_epsilon": 1e-05,
   "model_type": "gpt2",
   "n_ctx": 256,
-  "n_embd": 128,
-  "n_head": 4,
-  "n_inner": 128,
   "n_layer": 4,
   "n_positions": 1024,
   "reorder_and_upcast_attn": false,

 {
+  "_name_or_path": "gpt2",
   "activation_function": "gelu_new",
   "architectures": [
     "GPT2LMHeadModel"
   ],
+  "attn_pdrop": 0.1,
   "bos_token_id": 0,
   "embd_pdrop": 0.0,
   "eos_token_id": 0,
   "layer_norm_epsilon": 1e-05,
   "model_type": "gpt2",
   "n_ctx": 256,
+  "n_embd": 2048,
+  "n_head": 16,
+  "n_inner": 2048,
   "n_layer": 4,
   "n_positions": 1024,
   "reorder_and_upcast_attn": false,

pytorch_model.bin CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8f3de9c1853ae3e649aca4c6b9469b735c0c00a87a97e09425248e29c9308938
-size 7379877

 version https://git-lfs.github.com/spec/v1
+oid sha256:a6c569432850918cce7459bd9234c14e88b6e80fd0328bd479fa9fcb1135d96f
+size 432375793