Instructions to use zgcagi/ZGCM-1-7B-Midtrain-Staged-64K with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use zgcagi/ZGCM-1-7B-Midtrain-Staged-64K with Transformers:

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

pipe = pipeline("text-generation", model="zgcagi/ZGCM-1-7B-Midtrain-Staged-64K", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("zgcagi/ZGCM-1-7B-Midtrain-Staged-64K", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use zgcagi/ZGCM-1-7B-Midtrain-Staged-64K with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "zgcagi/ZGCM-1-7B-Midtrain-Staged-64K"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zgcagi/ZGCM-1-7B-Midtrain-Staged-64K",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/zgcagi/ZGCM-1-7B-Midtrain-Staged-64K

SGLang

How to use zgcagi/ZGCM-1-7B-Midtrain-Staged-64K 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 "zgcagi/ZGCM-1-7B-Midtrain-Staged-64K" \
    --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": "zgcagi/ZGCM-1-7B-Midtrain-Staged-64K",
		"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 "zgcagi/ZGCM-1-7B-Midtrain-Staged-64K" \
        --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": "zgcagi/ZGCM-1-7B-Midtrain-Staged-64K",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use zgcagi/ZGCM-1-7B-Midtrain-Staged-64K with Docker Model Runner:
```
docker model run hf.co/zgcagi/ZGCM-1-7B-Midtrain-Staged-64K
```

ZGCM-1-7B-Midtrain-Staged-64K

Completed 64K stage of the staged mid-training route, following the 32K stage.

Property	Value
Checkpoint step within its run	19206
Training context	65,536 tokens
Training tokens (mid-training cumulative, excluding pretraining)	approximately 422.16B
Precision	BF16
RoPE base	5,000,000
Stage completed	Yes

This is a base language-model checkpoint. It has not undergone instruction tuning. Use plain text inputs; the included tokenizer assets retain the original token IDs.

Loading

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "zgcagi/ZGCM-1-7B-Midtrain-Staged-64K"
tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
    model_id, trust_remote_code=True, torch_dtype=torch.bfloat16,
    device_map="auto", attn_implementation="eager",
)
inputs = tokenizer("The purpose of scientific research is", return_tensors="pt").to(model.device)
output = model.generate(**inputs, max_new_tokens=32, do_sample=False, use_cache=False)
print(tokenizer.decode(output[0], skip_special_tokens=True))

The repository includes the custom ZGCM Transformers implementation. The loading example uses its uncached reference generation path. Conversion and short forward/ generation checks used PyTorch 2.7 and Transformers 4.57.0.

Conversion validation

Exported with export_megatron_zgcm_to_hf.py. All 259 source model tensors round-trip exactly through the 382 exported tensors, preserving BF16 values, grouped Q/G/K/V layout, SwiGLU weights, and normalization parameters. Every exported shard is listed in model.safetensors.index.json and SHA256SUMS.

Local HF loading had no missing or unexpected tensors. A 160-token forward pass and short greedy generation passed. These checks do not constitute a native Megatron-versus-HF logits comparison or long-context benchmark validation.

See checkpoint_info.json for the precise checkpoint identity and validation scope.

Downloads last month: 12

Safetensors

Model size

7B params

Tensor type

BF16

Collection including zgcagi/ZGCM-1-7B-Midtrain-Staged-64K

ZGCM-1

Collection

8 items • Updated 1 day ago • 2