Instructions to use yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-4B")
model = PeftModel.from_pretrained(base_model, "yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B")

Transformers

How to use yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B with Transformers:

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

pipe = pipeline("text-generation", model="yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B

SGLang

How to use yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B 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 "yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B" \
    --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": "yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B",
		"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 "yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B" \
        --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": "yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B with Docker Model Runner:
```
docker model run hf.co/yihongLiu/COPSD-PolyMath-RUS-Qwen3-4B
```

COPSD-PolyMath-RUS-Qwen3-4B / trainer_state.json

yihongLiu

Upload COPSD PolyMath RU Qwen3-4B best checkpoint.

e04694d verified 1 day ago

raw

history blame contribute delete

3.66 kB

	{
	"best_global_step": null,
	"best_metric": null,
	"best_model_checkpoint": null,
	"epoch": 0.32,
	"eval_steps": 500,
	"global_step": 30,
	"is_hyper_param_search": false,
	"is_local_process_zero": true,
	"is_world_process_zero": true,
	"log_history": [
	{
	"epoch": 0.021333333333333333,
	"grad_norm": 0.13489535450935364,
	"learning_rate": 4.95e-06,
	"loss": 0.0012,
	"on_policy_loss": 0.0012,
	"step": 2
	},
	{
	"epoch": 0.042666666666666665,
	"grad_norm": 0.14185190200805664,
	"learning_rate": 4.85e-06,
	"loss": 0.0024,
	"on_policy_loss": 0.0024,
	"step": 4
	},
	{
	"epoch": 0.064,
	"grad_norm": 0.059850700199604034,
	"learning_rate": 4.75e-06,
	"loss": 0.0014,
	"on_policy_loss": 0.0014,
	"step": 6
	},
	{
	"epoch": 0.08533333333333333,
	"grad_norm": 0.08647190779447556,
	"learning_rate": 4.65e-06,
	"loss": 0.0003,
	"on_policy_loss": 0.0003,
	"step": 8
	},
	{
	"epoch": 0.10666666666666667,
	"grad_norm": 0.07889997214078903,
	"learning_rate": 4.5500000000000005e-06,
	"loss": 0.0002,
	"on_policy_loss": 0.0002,
	"step": 10
	},
	{
	"epoch": 0.128,
	"grad_norm": 0.12323067337274551,
	"learning_rate": 4.450000000000001e-06,
	"loss": -0.0005,
	"on_policy_loss": -0.0005,
	"step": 12
	},
	{
	"epoch": 0.14933333333333335,
	"grad_norm": 0.11012116074562073,
	"learning_rate": 4.350000000000001e-06,
	"loss": -0.0007,
	"on_policy_loss": -0.0007,
	"step": 14
	},
	{
	"epoch": 0.17066666666666666,
	"grad_norm": 0.20869049429893494,
	"learning_rate": 4.25e-06,
	"loss": -0.0014,
	"on_policy_loss": -0.0014,
	"step": 16
	},
	{
	"epoch": 0.192,
	"grad_norm": 0.0413811095058918,
	"learning_rate": 4.15e-06,
	"loss": -0.0019,
	"on_policy_loss": -0.0019,
	"step": 18
	},
	{
	"epoch": 0.21333333333333335,
	"grad_norm": 0.03595610707998276,
	"learning_rate": 4.05e-06,
	"loss": -0.0019,
	"on_policy_loss": -0.0019,
	"step": 20
	},
	{
	"epoch": 0.23466666666666666,
	"grad_norm": 0.03915691003203392,
	"learning_rate": 3.95e-06,
	"loss": -0.0028,
	"on_policy_loss": -0.0028,
	"step": 22
	},
	{
	"epoch": 0.256,
	"grad_norm": 0.0527644082903862,
	"learning_rate": 3.85e-06,
	"loss": -0.0027,
	"on_policy_loss": -0.0027,
	"step": 24
	},
	{
	"epoch": 0.2773333333333333,
	"grad_norm": 0.031156938523054123,
	"learning_rate": 3.7500000000000005e-06,
	"loss": -0.0035,
	"on_policy_loss": -0.0035,
	"step": 26
	},
	{
	"epoch": 0.2986666666666667,
	"grad_norm": 0.0408235602080822,
	"learning_rate": 3.65e-06,
	"loss": -0.0036,
	"on_policy_loss": -0.0036,
	"step": 28
	},
	{
	"epoch": 0.32,
	"grad_norm": 0.0373714454472065,
	"learning_rate": 3.5500000000000003e-06,
	"loss": -0.0038,
	"on_policy_loss": -0.0038,
	"step": 30
	}
	],
	"logging_steps": 2,
	"max_steps": 100,
	"num_input_tokens_seen": 0,
	"num_train_epochs": 2,
	"save_steps": 5,
	"stateful_callbacks": {
	"TrainerControl": {
	"args": {
	"should_epoch_stop": false,
	"should_evaluate": false,
	"should_log": false,
	"should_save": true,
	"should_training_stop": false
	},
	"attributes": {}
	}
	},
	"total_flos": 0.0,
	"train_batch_size": 4,
	"trial_name": null,
	"trial_params": null
	}