Instructions to use Kausp11/searchlm-sft_v4_inline_frombase8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Kausp11/searchlm-sft_v4_inline_frombase8 with Transformers:

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

pipe = pipeline("text-generation", model="Kausp11/searchlm-sft_v4_inline_frombase8")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Kausp11/searchlm-sft_v4_inline_frombase8")
model = AutoModelForCausalLM.from_pretrained("Kausp11/searchlm-sft_v4_inline_frombase8", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Kausp11/searchlm-sft_v4_inline_frombase8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Kausp11/searchlm-sft_v4_inline_frombase8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Kausp11/searchlm-sft_v4_inline_frombase8",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Kausp11/searchlm-sft_v4_inline_frombase8

SGLang

How to use Kausp11/searchlm-sft_v4_inline_frombase8 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 "Kausp11/searchlm-sft_v4_inline_frombase8" \
    --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": "Kausp11/searchlm-sft_v4_inline_frombase8",
		"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 "Kausp11/searchlm-sft_v4_inline_frombase8" \
        --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": "Kausp11/searchlm-sft_v4_inline_frombase8",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Kausp11/searchlm-sft_v4_inline_frombase8 with Docker Model Runner:
```
docker model run hf.co/Kausp11/searchlm-sft_v4_inline_frombase8
```

sft_v4_inline_frombase8

Role: sft_v4 inline-trace SFT retrofit on baseline_1p7B_v2_e8 (P1 retrofit)

Source: RCP search-lm/ckpts/sft_v4_inline_frombase8 (weights + tokenizer + trainer_state only; optimizer states and intermediate checkpoint-* dirs not uploaded). Pushed 2026-09-11. Numbers and protocol: Search-LLM RESULTS.md / EXPERIMENT_PLAN.md.

field	value
epoch	1.0
global_step	197
final_lr	8.644826107019888e-08
final_loss	1.8804
train_loss	1.9538462246735084
safetensors_bytes	3422794336

Downloads last month: 9

Safetensors

Model size

2B params

Tensor type

BF16