Instructions to use wefamm/aiAI_coder_V1.4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use wefamm/aiAI_coder_V1.4B with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("/root/autodl-tmp/Qwen3.5-4B-Thinking")
model = PeftModel.from_pretrained(base_model, "wefamm/aiAI_coder_V1.4B")

Transformers

How to use wefamm/aiAI_coder_V1.4B with Transformers:

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

pipe = pipeline("text-generation", model="wefamm/aiAI_coder_V1.4B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("wefamm/aiAI_coder_V1.4B", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use wefamm/aiAI_coder_V1.4B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "wefamm/aiAI_coder_V1.4B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "wefamm/aiAI_coder_V1.4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/wefamm/aiAI_coder_V1.4B

SGLang

How to use wefamm/aiAI_coder_V1.4B 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 "wefamm/aiAI_coder_V1.4B" \
    --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": "wefamm/aiAI_coder_V1.4B",
		"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 "wefamm/aiAI_coder_V1.4B" \
        --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": "wefamm/aiAI_coder_V1.4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use wefamm/aiAI_coder_V1.4B with Docker Model Runner:
```
docker model run hf.co/wefamm/aiAI_coder_V1.4B
```

aiAI_coder_V1.4B / training_args.yaml

wefamm

Upload folder using huggingface_hub

57ded2c verified 2 months ago

Raw History Blame Contribute Delete

1.02 kB

bf16: true
cutoff_len: 2048
dataset: coder4boptim
dataset_dir: /root/LLaMA-Factory/data
ddp_timeout: 180000000
do_train: true
double_quantization: true
enable_thinking: true
finetuning_type: lora
flash_attn: auto
freeze_multi_modal_projector: true
freeze_vision_tower: true
gradient_accumulation_steps: 8
image_max_pixels: 589824
image_min_pixels: 1024
include_num_input_tokens_seen: true
learning_rate: 5.0e-05
logging_steps: 5
lora_alpha: 16
lora_dropout: 0.01
lora_rank: 8
lora_target: all
loraplus_lr_ratio: 16
lr_scheduler_type: cosine
max_grad_norm: 1.0
max_samples: 6000
model_name_or_path: /root/autodl-tmp/Qwen3.5-4B-Thinking
num_train_epochs: 1.0
optim: adamw_torch
output_dir: saves/Qwen3.5-4B-Thinking/lora/aiAIL_coding_V1
packing: false
per_device_train_batch_size: 2
plot_loss: true
preprocessing_num_workers: 16
quantization_bit: 4
quantization_method: bnb
report_to: none
save_steps: 100
stage: sft
template: qwen3_5
trust_remote_code: true
video_max_pixels: 65536
video_min_pixels: 256
warmup_steps: 0

	bf16: true
	cutoff_len: 2048
	dataset: coder4boptim
	dataset_dir: /root/LLaMA-Factory/data
	ddp_timeout: 180000000
	do_train: true
	double_quantization: true
	enable_thinking: true
	finetuning_type: lora
	flash_attn: auto
	freeze_multi_modal_projector: true
	freeze_vision_tower: true
	gradient_accumulation_steps: 8
	image_max_pixels: 589824
	image_min_pixels: 1024
	include_num_input_tokens_seen: true
	learning_rate: 5.0e-05
	logging_steps: 5
	lora_alpha: 16
	lora_dropout: 0.01
	lora_rank: 8
	lora_target: all
	loraplus_lr_ratio: 16
	lr_scheduler_type: cosine
	max_grad_norm: 1.0
	max_samples: 6000
	model_name_or_path: /root/autodl-tmp/Qwen3.5-4B-Thinking
	num_train_epochs: 1.0
	optim: adamw_torch
	output_dir: saves/Qwen3.5-4B-Thinking/lora/aiAIL_coding_V1
	packing: false
	per_device_train_batch_size: 2
	plot_loss: true
	preprocessing_num_workers: 16
	quantization_bit: 4
	quantization_method: bnb
	report_to: none
	save_steps: 100
	stage: sft
	template: qwen3_5
	trust_remote_code: true
	video_max_pixels: 65536
	video_min_pixels: 256
	warmup_steps: 0