Instructions to use thuml/Thoth-30B-A3B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use thuml/Thoth-30B-A3B with Transformers:

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

pipe = pipeline("text-generation", model="thuml/Thoth-30B-A3B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("thuml/Thoth-30B-A3B")
model = AutoModelForCausalLM.from_pretrained("thuml/Thoth-30B-A3B")
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 thuml/Thoth-30B-A3B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "thuml/Thoth-30B-A3B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "thuml/Thoth-30B-A3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/thuml/Thoth-30B-A3B

SGLang

How to use thuml/Thoth-30B-A3B 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 "thuml/Thoth-30B-A3B" \
    --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": "thuml/Thoth-30B-A3B",
		"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 "thuml/Thoth-30B-A3B" \
        --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": "thuml/Thoth-30B-A3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use thuml/Thoth-30B-A3B with Docker Model Runner:
```
docker model run hf.co/thuml/Thoth-30B-A3B
```

exLinnn commited on Mar 1

Commit

ef6da2c

verified ·

1 Parent(s): e0febc4

init readme

Browse files

Files changed (2) hide show

README.md +26 -3
model.safetensors.index.json +1 -2

README.md CHANGED Viewed

@@ -1,3 +1,26 @@
----
-license: apache-2.0
----

+---
+license: apache-2.0
+base_model: Qwen/Qwen3-30B-A3B-Instruct-2507
+pipeline_tag: text-generation
+tags:
+  - large language models
+  - qwen3
+  - time series
+  - mid-training
+  - question-answering
+arxiv:
+library_name: transformers
+---
+# Thoth
+Thoth introduces the first family of mid-trained LLMs specifically enhanced for time series understanding through a task- and domain-agnostic mid-training stage.
+See https://github.com/thuml/Thoth for examples for using this model.
+## Citation
+If you find our work useful, please cite our paper as:
+```
+```

model.safetensors.index.json CHANGED Viewed

@@ -1,6 +1,5 @@
 {
   "metadata": {
-    "total_parameters": 210944,
     "total_size": 61064245248
   },
   "weight_map": {
@@ -18872,4 +18871,4 @@
     "model.layers.9.self_attn.v_proj.weight": "model-00003-of-00013.safetensors",
     "model.norm.weight": "model-00013-of-00013.safetensors"
   }
-}

 {
   "metadata": {
     "total_size": 61064245248
   },
   "weight_map": {
     "model.layers.9.self_attn.v_proj.weight": "model-00003-of-00013.safetensors",
     "model.norm.weight": "model-00013-of-00013.safetensors"
   }
+}