Instructions to use codellama/CodeLlama-7b-Python-hf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use codellama/CodeLlama-7b-Python-hf with Transformers:

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

pipe = pipeline("text-generation", model="codellama/CodeLlama-7b-Python-hf")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("codellama/CodeLlama-7b-Python-hf")
model = AutoModelForCausalLM.from_pretrained("codellama/CodeLlama-7b-Python-hf")

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use codellama/CodeLlama-7b-Python-hf with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "codellama/CodeLlama-7b-Python-hf"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "codellama/CodeLlama-7b-Python-hf",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/codellama/CodeLlama-7b-Python-hf

SGLang

How to use codellama/CodeLlama-7b-Python-hf 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 "codellama/CodeLlama-7b-Python-hf" \
    --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": "codellama/CodeLlama-7b-Python-hf",
		"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 "codellama/CodeLlama-7b-Python-hf" \
        --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": "codellama/CodeLlama-7b-Python-hf",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use codellama/CodeLlama-7b-Python-hf with Docker Model Runner:
```
docker model run hf.co/codellama/CodeLlama-7b-Python-hf
```

achalagarwal commited on Nov 14, 2023

Commit

b46bda9

1 Parent(s): 7ee7b6b

Update tokenizer_config.json

Browse files

Following https://huggingface.co/codellama/CodeLlama-34b-hf/discussions/9 to support usage with ctranslate2
cc: @ctranslate2-4you

Files changed (1) hide show

tokenizer_config.json +7 -0

tokenizer_config.json CHANGED Viewed

@@ -1,6 +1,7 @@
 {
   "add_bos_token": true,
   "add_eos_token": false,
   "bos_token": {
     "__type": "AddedToken",
     "content": "<s>",
@@ -9,6 +10,8 @@
     "rstrip": false,
     "single_word": false
   },
   "clean_up_tokenization_spaces": false,
   "eos_token": {
     "__type": "AddedToken",
@@ -19,9 +22,13 @@
     "single_word": false
   },
   "legacy": null,
   "model_max_length": 1000000000000000019884624838656,
   "pad_token": null,
   "sp_model_kwargs": {},
   "tokenizer_class": "CodeLlamaTokenizer",
   "unk_token": {
     "__type": "AddedToken",

 {
   "add_bos_token": true,
   "add_eos_token": false,
+  "additional_special_tokens": [],
   "bos_token": {
     "__type": "AddedToken",
     "content": "<s>",
     "rstrip": false,
     "single_word": false
   },
+  "eot_token": null,
+  "fill_token": "<FILL_ME>",
   "clean_up_tokenization_spaces": false,
   "eos_token": {
     "__type": "AddedToken",
     "single_word": false
   },
   "legacy": null,
+  "middle_token": null,
   "model_max_length": 1000000000000000019884624838656,
   "pad_token": null,
+  "prefix_token": null,
   "sp_model_kwargs": {},
+  "suffix_first": false,
+  "suffix_token": null,
   "tokenizer_class": "CodeLlamaTokenizer",
   "unk_token": {
     "__type": "AddedToken",