Instructions to use LayerEight/ISP-Qwen2.5-7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use LayerEight/ISP-Qwen2.5-7B with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-7B-Instruct")
model = PeftModel.from_pretrained(base_model, "LayerEight/ISP-Qwen2.5-7B")

Transformers

How to use LayerEight/ISP-Qwen2.5-7B with Transformers:

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

pipe = pipeline("text-generation", model="LayerEight/ISP-Qwen2.5-7B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("LayerEight/ISP-Qwen2.5-7B", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use LayerEight/ISP-Qwen2.5-7B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "LayerEight/ISP-Qwen2.5-7B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "LayerEight/ISP-Qwen2.5-7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/LayerEight/ISP-Qwen2.5-7B

SGLang

How to use LayerEight/ISP-Qwen2.5-7B 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 "LayerEight/ISP-Qwen2.5-7B" \
    --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": "LayerEight/ISP-Qwen2.5-7B",
		"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 "LayerEight/ISP-Qwen2.5-7B" \
        --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": "LayerEight/ISP-Qwen2.5-7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use LayerEight/ISP-Qwen2.5-7B with Docker Model Runner:
```
docker model run hf.co/LayerEight/ISP-Qwen2.5-7B
```

ISP-Qwen2.5-7B

File size: 2,247 Bytes

4e41e5d

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 148,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.8343619173765182,
      "epoch": 0.16891891891891891,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0001914140188963917,
      "loss": 1.8805531311035155,
      "mean_token_accuracy": 0.5832486696541309,
      "num_tokens": 135904.0,
      "step": 25
    },
    {
      "entropy": 1.6487643253803252,
      "epoch": 0.33783783783783783,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.00015680647467311557,
      "loss": 1.6185850524902343,
      "mean_token_accuracy": 0.6362091641128064,
      "num_tokens": 266483.0,
      "step": 50
    },
    {
      "entropy": 1.3490780843421817,
      "epoch": 0.5067567567567568,
      "grad_norm": 0.140625,
      "learning_rate": 0.00010548953398643275,
      "loss": 1.2554483032226562,
      "mean_token_accuracy": 0.7003054420650006,
      "num_tokens": 394707.0,
      "step": 75
    },
    {
      "entropy": 1.3766383936069906,
      "epoch": 0.6756756756756757,
      "grad_norm": 0.09814453125,
      "learning_rate": 5.2557868916543994e-05,
      "loss": 1.3398623657226563,
      "mean_token_accuracy": 0.6991272465884686,
      "num_tokens": 525512.0,
      "step": 100
    },
    {
      "entropy": 1.243874678350985,
      "epoch": 0.8445945945945946,
      "grad_norm": 0.087890625,
      "learning_rate": 1.3581116238315195e-05,
      "loss": 1.1492684173583985,
      "mean_token_accuracy": 0.7265925395488739,
      "num_tokens": 651191.0,
      "step": 125
    }
  ],
  "logging_steps": 25,
  "max_steps": 148,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.2900897797103616e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}