Instructions to use youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned with Transformers:

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

pipe = pipeline("text-generation", model="youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned")
model = AutoModelForCausalLM.from_pretrained("youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned")
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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned

SGLang

How to use youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned 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 "youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned" \
    --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": "youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned",
		"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 "youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned" \
        --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": "youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned with Docker Model Runner:
```
docker model run hf.co/youssefbelghmi/MNLP_M3_mcqa_model_PreFineTuned
```

MNLP_M3_mcqa_model_PreFineTuned / checkpoint-904 /trainer_state.json

youssefbelghmi

Upload folder using huggingface_hub

2e6205c verified 12 months ago

raw

history blame contribute delete

4.99 kB

	{
	"best_global_step": null,
	"best_metric": null,
	"best_model_checkpoint": null,
	"epoch": 2.0,
	"eval_steps": 100,
	"global_step": 904,
	"is_hyper_param_search": false,
	"is_local_process_zero": true,
	"is_world_process_zero": true,
	"log_history": [
	{
	"epoch": 0.22136137244050913,
	"grad_norm": 1.6171875,
	"learning_rate": 0.00017809734513274337,
	"loss": 1.2949,
	"step": 100
	},
	{
	"epoch": 0.22136137244050913,
	"eval_loss": 1.2047334909439087,
	"eval_mean_token_accuracy": 0.7032503170787163,
	"eval_num_tokens": 1635208.0,
	"eval_runtime": 55.1857,
	"eval_samples_per_second": 23.031,
	"eval_steps_per_second": 2.881,
	"step": 100
	},
	{
	"epoch": 0.44272274488101826,
	"grad_norm": 1.5234375,
	"learning_rate": 0.00015597345132743363,
	"loss": 1.187,
	"step": 200
	},
	{
	"epoch": 0.44272274488101826,
	"eval_loss": 1.1497339010238647,
	"eval_mean_token_accuracy": 0.7146404561756542,
	"eval_num_tokens": 3271888.0,
	"eval_runtime": 54.9525,
	"eval_samples_per_second": 23.129,
	"eval_steps_per_second": 2.893,
	"step": 200
	},
	{
	"epoch": 0.6640841173215274,
	"grad_norm": 1.4140625,
	"learning_rate": 0.00013384955752212388,
	"loss": 1.1394,
	"step": 300
	},
	{
	"epoch": 0.6640841173215274,
	"eval_loss": 1.1108253002166748,
	"eval_mean_token_accuracy": 0.7227962376936427,
	"eval_num_tokens": 4908096.0,
	"eval_runtime": 54.9495,
	"eval_samples_per_second": 23.13,
	"eval_steps_per_second": 2.894,
	"step": 300
	},
	{
	"epoch": 0.8854454897620365,
	"grad_norm": 1.484375,
	"learning_rate": 0.00011172566371681417,
	"loss": 1.0831,
	"step": 400
	},
	{
	"epoch": 0.8854454897620365,
	"eval_loss": 1.0701380968093872,
	"eval_mean_token_accuracy": 0.7318694584024777,
	"eval_num_tokens": 6544993.0,
	"eval_runtime": 54.9634,
	"eval_samples_per_second": 23.124,
	"eval_steps_per_second": 2.893,
	"step": 400
	},
	{
	"epoch": 1.1062534587714443,
	"grad_norm": 1.296875,
	"learning_rate": 8.960176991150443e-05,
	"loss": 0.8872,
	"step": 500
	},
	{
	"epoch": 1.1062534587714443,
	"eval_loss": 1.084444284439087,
	"eval_mean_token_accuracy": 0.735802007921087,
	"eval_num_tokens": 8174028.0,
	"eval_runtime": 54.9918,
	"eval_samples_per_second": 23.113,
	"eval_steps_per_second": 2.891,
	"step": 500
	},
	{
	"epoch": 1.3276148312119536,
	"grad_norm": 1.4140625,
	"learning_rate": 6.747787610619469e-05,
	"loss": 0.6883,
	"step": 600
	},
	{
	"epoch": 1.3276148312119536,
	"eval_loss": 1.0781114101409912,
	"eval_mean_token_accuracy": 0.7385061788109114,
	"eval_num_tokens": 9808348.0,
	"eval_runtime": 54.9555,
	"eval_samples_per_second": 23.128,
	"eval_steps_per_second": 2.893,
	"step": 600
	},
	{
	"epoch": 1.5489762036524626,
	"grad_norm": 1.375,
	"learning_rate": 4.535398230088496e-05,
	"loss": 0.6754,
	"step": 700
	},
	{
	"epoch": 1.5489762036524626,
	"eval_loss": 1.0621684789657593,
	"eval_mean_token_accuracy": 0.741633175304101,
	"eval_num_tokens": 11445630.0,
	"eval_runtime": 54.946,
	"eval_samples_per_second": 23.132,
	"eval_steps_per_second": 2.894,
	"step": 700
	},
	{
	"epoch": 1.7703375760929718,
	"grad_norm": 1.328125,
	"learning_rate": 2.3230088495575223e-05,
	"loss": 0.6565,
	"step": 800
	},
	{
	"epoch": 1.7703375760929718,
	"eval_loss": 1.0559310913085938,
	"eval_mean_token_accuracy": 0.7433326304333765,
	"eval_num_tokens": 13082315.0,
	"eval_runtime": 54.9745,
	"eval_samples_per_second": 23.12,
	"eval_steps_per_second": 2.892,
	"step": 800
	},
	{
	"epoch": 1.991698948533481,
	"grad_norm": 1.375,
	"learning_rate": 1.1061946902654867e-06,
	"loss": 0.6583,
	"step": 900
	},
	{
	"epoch": 1.991698948533481,
	"eval_loss": 1.055925726890564,
	"eval_mean_token_accuracy": 0.7435795432366665,
	"eval_num_tokens": 14718777.0,
	"eval_runtime": 54.9647,
	"eval_samples_per_second": 23.124,
	"eval_steps_per_second": 2.893,
	"step": 900
	}
	],
	"logging_steps": 100,
	"max_steps": 904,
	"num_input_tokens_seen": 0,
	"num_train_epochs": 2,
	"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": 3.91050531176448e+16,
	"train_batch_size": 4,
	"trial_name": null,
	"trial_params": null
	}