Instructions to use harims95/LoopLM-135M-naive-sft with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use harims95/LoopLM-135M-naive-sft with Transformers:

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

pipe = pipeline("text-generation", model="harims95/LoopLM-135M-naive-sft", trust_remote_code=True, device_map="auto")

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("harims95/LoopLM-135M-naive-sft", trust_remote_code=True, dtype="auto", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use harims95/LoopLM-135M-naive-sft with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "harims95/LoopLM-135M-naive-sft"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "harims95/LoopLM-135M-naive-sft",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/harims95/LoopLM-135M-naive-sft

SGLang

How to use harims95/LoopLM-135M-naive-sft 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 "harims95/LoopLM-135M-naive-sft" \
    --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": "harims95/LoopLM-135M-naive-sft",
		"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 "harims95/LoopLM-135M-naive-sft" \
        --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": "harims95/LoopLM-135M-naive-sft",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use harims95/LoopLM-135M-naive-sft with Docker Model Runner:
```
docker model run hf.co/harims95/LoopLM-135M-naive-sft
```

LoopLM-135M-naive-sft / config.json

harims95

fix: checkpoint-4878 with correctly tied lm_head weights

f49d29f verified 22 days ago

Raw

History Blame Contribute Delete

617 Bytes

	{
	"architectures": [
	"LoopLMForCausalLM"
	],
	"auto_map": {
	"AutoConfig": "configuration_looplm.LoopLMConfig",
	"AutoModelForCausalLM": "modeling_looplm.LoopLMForCausalLM"
	},
	"d_model": 1024,
	"dense_ffn": 2816,
	"dtype": "float32",
	"final_z_loss_coef": 0.0001,
	"head_dim": 64,
	"init_std": 0.02,
	"model_type": "looplm",
	"mu_rec": 6,
	"n_coda": 2,
	"n_kv_heads": 8,
	"n_prelude": 4,
	"n_q_heads": 16,
	"qk_norm": true,
	"rope_theta": 10000.0,
	"tie_embeddings": true,
	"transformers_version": "5.12.1",
	"use_a_matrix": false,
	"use_input_norm": false,
	"vocab_size": 50304
	}