Instructions to use eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2 with Transformers:

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

pipe = pipeline("text-generation", model="eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2")
model = AutoModelForCausalLM.from_pretrained("eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2")
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 Settings

vLLM

How to use eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2

SGLang

How to use eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2 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 "eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2" \
    --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": "eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2",
		"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 "eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2" \
        --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": "eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2 with Docker Model Runner:
```
docker model run hf.co/eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2
```

OpenThinkerAgent-32B-AWQ-Int4-Terminus2 / quantization_run.json

eewer

Upload OpenThinkerAgent 32B AWQ int4 Terminus-2 quantization

f1f1905 verified 7 days ago

Raw

History Blame Contribute Delete

1.74 kB

	{
	"apply_clip": true,
	"autoawq_version": "0.2.9",
	"bits": 4,
	"calibration_max_chars": null,
	"calibration_max_tokens": 2048,
	"calibration_min_chars": 256,
	"calibration_samples": 128,
	"calibration_seed": 7,
	"command": [
	"terminal_agent_demo/scripts/quantize_openthinker_awq.py",
	"--output-dir",
	"quantization/models/OpenThinkerAgent-32B-AWQ-Int4-Terminus2",
	"--repo-id",
	"eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2",
	"--samples",
	"128",
	"--max-tokens",
	"2048",
	"--max-calib-seq-len",
	"2048",
	"--n-parallel-calib-samples",
	"1",
	"--max-chunk-memory-mb",
	"256",
	"--group-size",
	"128",
	"--device-map",
	"auto",
	"--max-shard-size",
	"5GB"
	],
	"cuda_version": "12.8",
	"device_map": "auto",
	"duo_scaling": true,
	"elapsed_seconds": 12111.596,
	"finished_at_utc": "2026-06-25T01:31:16.596143+00:00",
	"group_size": 128,
	"max_calib_seq_len": 2048,
	"max_chunk_memory_mb": 256,
	"modules_to_not_convert": [
	"lm_head"
	],
	"n_parallel_calib_samples": 1,
	"output_dir": "/wbl-fast/usrs/ee/clean-20260619/worktrees/terminal-agent-rl-opd-qwen3-4b/quantization/models/OpenThinkerAgent-32B-AWQ-Int4-Terminus2",
	"repo_id": "eewer/OpenThinkerAgent-32B-AWQ-Int4-Terminus2",
	"source_dataset": "/wbl-fast/usrs/ee/clean-20260619/terminal-agent-rl/areal_runs/terminal-agent-demo/data/skill_based_medium.even.terminus2.slime_messages.jsonl",
	"source_model": "open-thoughts/OpenThinkerAgent-32B",
	"source_revision": "65d8a62b87c8d3d34bc45108a7ad87635318db9f",
	"started_at_utc": "2026-06-24T22:09:24.999975+00:00",
	"torch_version": "2.11.0+cu128",
	"transformers_version": "5.12.1",
	"version": "GEMM",
	"zero_point": true
	}