Instructions to use i-be-snek/moe_tur_multi_batch_8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use i-be-snek/moe_tur_multi_batch_8 with Transformers:

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

pipe = pipeline("text-generation", model="i-be-snek/moe_tur_multi_batch_8")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("i-be-snek/moe_tur_multi_batch_8")
model = AutoModelForCausalLM.from_pretrained("i-be-snek/moe_tur_multi_batch_8")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use i-be-snek/moe_tur_multi_batch_8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "i-be-snek/moe_tur_multi_batch_8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "i-be-snek/moe_tur_multi_batch_8",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/i-be-snek/moe_tur_multi_batch_8

SGLang

How to use i-be-snek/moe_tur_multi_batch_8 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 "i-be-snek/moe_tur_multi_batch_8" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "i-be-snek/moe_tur_multi_batch_8",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "i-be-snek/moe_tur_multi_batch_8" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "i-be-snek/moe_tur_multi_batch_8",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use i-be-snek/moe_tur_multi_batch_8 with Docker Model Runner:
```
docker model run hf.co/i-be-snek/moe_tur_multi_batch_8
```

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

moe_tur_multi_batch_8

This model is a fine-tuned version of on the arrow dataset. It achieves the following results on the evaluation set:

Loss: 5.9074

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0001
train_batch_size: 2
eval_batch_size: 32
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 8
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-06 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 38244
training_steps: 382446
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0	0	10.9789
7.4445	0.2615	10000	7.3980
5.6546	0.5229	20000	5.6101
5.036	0.7844	30000	5.0138
4.6238	1.0459	40000	4.6896
4.445	1.3074	50000	4.4645
4.2684	1.5689	60000	4.3137
4.2097	1.8303	70000	4.2050
3.9034	2.0918	80000	4.1332
3.8426	2.3533	90000	4.0845
3.8379	2.6148	100000	4.0366
3.8142	2.8763	110000	3.9902
3.4415	3.1377	120000	4.0038
3.5026	3.3992	130000	3.9916
3.5018	3.6607	140000	3.9655
3.5207	3.9222	150000	3.9371
3.5128	4.0000	152976	3.9287
2.9641	4.1837	160000	4.0548
3.038	4.4451	170000	4.0626
3.1051	4.7066	180000	4.0439
3.1606	4.9681	190000	4.0237
2.4678	5.2296	200000	4.2661
2.5802	5.4910	210000	4.2925
2.6698	5.7525	220000	4.2789
1.8528	6.0140	230000	4.3869
1.9577	6.2755	240000	4.6120
2.0102	6.5370	250000	4.6567
2.097	6.7984	260000	4.6590
2.1054	7.0000	267708	4.6630
1.3511	7.0599	270000	4.8671
1.3987	7.3214	280000	5.0544
1.5008	7.5829	290000	5.1165
1.5377	7.8444	300000	5.1459
0.9324	8.1058	310000	5.3730
0.9891	8.3673	320000	5.5069
1.0269	8.6288	330000	5.5789
1.0243	8.8903	340000	5.6237
0.6486	9.1518	350000	5.7925
0.6572	9.4132	360000	5.8604
0.6697	9.6747	370000	5.8925
0.6738	9.9362	380000	5.9074

Framework versions

Transformers 4.51.0
Pytorch 2.7.0+cu126
Datasets 3.6.0
Tokenizers 0.21.1

Downloads last month: -

Safetensors

Model size

0.5B params

Tensor type

F32