Instructions to use SurjoLabs/Blaze-Title with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SurjoLabs/Blaze-Title with Transformers:

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

pipe = pipeline("text-generation", model="SurjoLabs/Blaze-Title", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("SurjoLabs/Blaze-Title", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use SurjoLabs/Blaze-Title with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/SurjoLabs/Blaze-Title

SGLang

How to use SurjoLabs/Blaze-Title 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 "SurjoLabs/Blaze-Title" \
    --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": "SurjoLabs/Blaze-Title",
		"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 "SurjoLabs/Blaze-Title" \
        --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": "SurjoLabs/Blaze-Title",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use SurjoLabs/Blaze-Title with Docker Model Runner:
```
docker model run hf.co/SurjoLabs/Blaze-Title
```

Blaze-Title / configuration_blaze.py

spitfire4794

Blaze-Title — FFT on Qyrou 115K + Wild 130K (70/30, wild deduped), lr=0.0003, 1 epoch, seq 2048, assistant-only loss, compile default

7b59061 verified 27 days ago

Raw History Blame Contribute Delete

1.2 kB


	from transformers import LlamaConfig

	class BlazeConfig(LlamaConfig):
	model_type = "blaze"
	def __init__(self, *args, xsa_projection=True, rope_theta=10000.0, attention_bias=False,
	prelude_layers=1, recurrent_layers=12, coda_layers=1,
	recurrent_passes=2,
	gradient_checkpointing=False, use_flash_attn=True, **kwargs):
	kwargs["num_hidden_layers"] = prelude_layers + recurrent_layers + coda_layers
	kwargs.setdefault("use_cache", False)
	super().__init__(args, rope_theta=rope_theta, attention_bias=attention_bias, *kwargs)
	self.xsa_projection = xsa_projection
	self.rope_theta = rope_theta
	self.attention_bias = attention_bias
	self.prelude_layers = prelude_layers
	self.recurrent_layers = recurrent_layers
	self.coda_layers = coda_layers
	self.recurrent_passes = recurrent_passes
	self.gradient_checkpointing = gradient_checkpointing
	self.use_flash_attn = use_flash_attn
	if not hasattr(self, 'rope_parameters') or self.rope_parameters is None:
	self.rope_parameters = {"rope_type": "default", "factor": 1.0, "rope_theta": rope_theta}