Instructions to use whitecircle/GLM-4.7-Flash-Coder with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use whitecircle/GLM-4.7-Flash-Coder with Transformers:

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

pipe = pipeline("text-generation", model="whitecircle/GLM-4.7-Flash-Coder")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

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

tokenizer = AutoTokenizer.from_pretrained("whitecircle/GLM-4.7-Flash-Coder")
model = AutoModelForCausalLM.from_pretrained("whitecircle/GLM-4.7-Flash-Coder", device_map="auto")
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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use whitecircle/GLM-4.7-Flash-Coder with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "whitecircle/GLM-4.7-Flash-Coder"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "whitecircle/GLM-4.7-Flash-Coder",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/whitecircle/GLM-4.7-Flash-Coder

SGLang

How to use whitecircle/GLM-4.7-Flash-Coder 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 "whitecircle/GLM-4.7-Flash-Coder" \
    --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": "whitecircle/GLM-4.7-Flash-Coder",
		"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 "whitecircle/GLM-4.7-Flash-Coder" \
        --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": "whitecircle/GLM-4.7-Flash-Coder",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use whitecircle/GLM-4.7-Flash-Coder with Docker Model Runner:
```
docker model run hf.co/whitecircle/GLM-4.7-Flash-Coder
```

olegsmirnov commited on Jul 16

Commit

3d68b8d

verified ·

1 Parent(s): 4364941

Unfold Halo config section, remove leading br in details blocks

Browse files

Files changed (1) hide show

README.md +1 -7

README.md CHANGED Viewed

@@ -50,8 +50,7 @@ We trained Coder with Supervised Fine-Tuning (SFT) using [Halo](https://github.c
 - Learning rate scheduling: Cosine, with linear warm-up
 - Training time: 3 hours on 16 B300 GPUs
-<details open>
-<summary><h3>Halo 😇 is driven entirely by a single YAML config</h3></summary>
 ```yaml
 model_name_or_path: zai-org/GLM-4.7-Flash
@@ -98,8 +97,6 @@ run_name: glm-4.7-flash-coder
 enable_efficiency_metrics: true
 ```
-</details>
 ## Emergent skills
 The model picked up several of the teacher’s skills and patterns during fine-tuning:
@@ -200,7 +197,6 @@ Inference costs are calculated from OpenRouter pricing tables. We report the the
 <details>
 <summary>OpenRouter providers per model</summary>
-<br>
 | model | provider | input $/Mtok | output $/Mtok | cache_read $/Mtok | cache_write $/Mtok |
 | --- | --- | --- | --- | --- | --- |
@@ -231,7 +227,6 @@ Inference costs are calculated from OpenRouter pricing tables. We report the the
 <details>
 <summary>Prompt template</summary>
-<br>
 ```markdown
 You are solving a real GitHub issue in the `{repo}` repository. The repo is already cloned and set up in the current working directory.
@@ -260,7 +255,6 @@ The following interfaces are expected by the test suite. Pay close attention to
 <details>
 <summary>Inference with SGLang 0.5.14, CUDA 13</summary>
-<br>
 ```bash
 python3 -m sglang_router.launch_server \

 - Learning rate scheduling: Cosine, with linear warm-up
 - Training time: 3 hours on 16 B300 GPUs
+### Halo 😇 is driven entirely by a single YAML config
 ```yaml
 model_name_or_path: zai-org/GLM-4.7-Flash
 enable_efficiency_metrics: true
 ```
 ## Emergent skills
 The model picked up several of the teacher’s skills and patterns during fine-tuning:
 <details>
 <summary>OpenRouter providers per model</summary>
 | model | provider | input $/Mtok | output $/Mtok | cache_read $/Mtok | cache_write $/Mtok |
 | --- | --- | --- | --- | --- | --- |
 <details>
 <summary>Prompt template</summary>
 ```markdown
 You are solving a real GitHub issue in the `{repo}` repository. The repo is already cloned and set up in the current working directory.
 <details>
 <summary>Inference with SGLang 0.5.14, CUDA 13</summary>
 ```bash
 python3 -m sglang_router.launch_server \