Instructions to use manu/CroissantLLM-PII-filter with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use manu/CroissantLLM-PII-filter with Transformers:

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

pipe = pipeline("text-generation", model="manu/CroissantLLM-PII-filter")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("manu/CroissantLLM-PII-filter")
model = AutoModelForCausalLM.from_pretrained("manu/CroissantLLM-PII-filter")
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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use manu/CroissantLLM-PII-filter with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "manu/CroissantLLM-PII-filter"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "manu/CroissantLLM-PII-filter",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/manu/CroissantLLM-PII-filter

SGLang

How to use manu/CroissantLLM-PII-filter 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 "manu/CroissantLLM-PII-filter" \
    --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": "manu/CroissantLLM-PII-filter",
		"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 "manu/CroissantLLM-PII-filter" \
        --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": "manu/CroissantLLM-PII-filter",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use manu/CroissantLLM-PII-filter with Docker Model Runner:
```
docker model run hf.co/manu/CroissantLLM-PII-filter
```

See axolotl config

axolotl version: 0.4.0

base_model: croissantllm/CroissantLLMBase                                                                                                                                                                   
model_type: LlamaForCausalLM                                                                                                                                                                                
tokenizer_type: LlamaTokenizerFast                                                                                                                                                                              
is_llama_derived_model: true                                                                                                                                                                                

special_tokens:
  bos_token: "<s>"
  eos_token: "</s>"
  unk_token: "<unk>"

tokens:
  - "<|im_start|>"
  - "<|im_end|>"
                                                                                                                                                                                                            
load_in_8bit: false                                                                                                                                                                                         
load_in_4bit: false                                                                                                                                                                                         
strict: false                                                                                                                                                                                               
                                                                                                                                                                                                            
datasets:                                                                                                                                                                                                   
  - path: manu/pii-masking
    split: train                                                                                                                                                              
    type: sharegpt

chat_template: "chatml"
default_system_message: "Rewrite the sentence to remove the PII."

dataset_prepared_path: last_pii  
val_set_size: 0.05                       
output_dir: ./out_pii
                                                                                                                                                                                                            
sequence_len: 2048                                                                                                                                                                                          
sample_packing: false                                                                                                                                                                                       
pad_to_sequence_len: false                                                                                                                                                                                   
                                                                                                                                                                                                            
adapter:                                                                                                                                                                                                    
lora_model_dir:                                                                                                                                                                                             
lora_r:                                                                                                                                                                                                     
lora_alpha:                                                                                                                                                                                                 
lora_dropout:
lora_target_linear:
lora_fan_in_fan_out:

wandb_project:
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:

gradient_accumulation_steps: 2
micro_batch_size: 16
num_epochs: 1
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.00003

train_on_inputs: false
group_by_length: false
bf16: auto
fp16: false
tf32: true

gradient_checkpointing: true
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true
flash_attn_cross_entropy: false
flash_attn_rms_norm: true
flash_attn_fuse_qkv: false
flash_attn_fuse_mlp: true

warmup_steps: 100
evals_per_epoch: 4
eval_table_size:
saves_per_epoch: 1
debug:
deepspeed: #deepspeed_configs/zero2.json # multi-gpu only
weight_decay: 0.05
fsdp:
fsdp_config:

out_pii

This model is a fine-tuned version of croissantllm/CroissantLLMBase on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.0082

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 3e-05
train_batch_size: 16
eval_batch_size: 16
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 100
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
6.0535	0.0	1	6.2368
0.0107	0.25	783	0.0137
0.013	0.5	1566	0.0098
0.0077	0.75	2349	0.0082

Framework versions

Transformers 4.37.2
Pytorch 2.1.2+cu121
Datasets 2.16.1
Tokenizers 0.15.0

Downloads last month: 5

Model tree for manu/CroissantLLM-PII-filter

Base model

croissantllm/CroissantLLMBase

Finetuned

(5)

this model