Instructions to use GItaf/roberta-base-roberta-base-TF-weight1-epoch15 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use GItaf/roberta-base-roberta-base-TF-weight1-epoch15 with Transformers:

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

pipe = pipeline("text-generation", model="GItaf/roberta-base-roberta-base-TF-weight1-epoch15", device_map="auto")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("GItaf/roberta-base-roberta-base-TF-weight1-epoch15")
model = AutoModelForCausalLM.from_pretrained("GItaf/roberta-base-roberta-base-TF-weight1-epoch15", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use GItaf/roberta-base-roberta-base-TF-weight1-epoch15 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "GItaf/roberta-base-roberta-base-TF-weight1-epoch15"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "GItaf/roberta-base-roberta-base-TF-weight1-epoch15",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/GItaf/roberta-base-roberta-base-TF-weight1-epoch15

SGLang

How to use GItaf/roberta-base-roberta-base-TF-weight1-epoch15 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 "GItaf/roberta-base-roberta-base-TF-weight1-epoch15" \
    --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": "GItaf/roberta-base-roberta-base-TF-weight1-epoch15",
		"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 "GItaf/roberta-base-roberta-base-TF-weight1-epoch15" \
        --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": "GItaf/roberta-base-roberta-base-TF-weight1-epoch15",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use GItaf/roberta-base-roberta-base-TF-weight1-epoch15 with Docker Model Runner:
```
docker model run hf.co/GItaf/roberta-base-roberta-base-TF-weight1-epoch15
```

roberta-base-roberta-base-TF-weight1-epoch15

This model is a fine-tuned version of on the None dataset. It achieves the following results on the evaluation set:

Loss: 4.8322
Cls loss: 0.6900
Lm loss: 4.1423
Cls Accuracy: 0.5401
Cls F1: 0.3788
Cls Precision: 0.2917
Cls Recall: 0.5401
Perplexity: 62.95

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: 2e-05
train_batch_size: 2
eval_batch_size: 2
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Cls loss	Lm loss	Cls Accuracy	Cls F1	Cls Precision	Cls Recall	Perplexity
5.3158	1.0	3470	4.9858	0.6910	4.2949	0.5401	0.3788	0.2917	0.5401	73.32
4.9772	2.0	6940	4.8876	0.6956	4.1920	0.4599	0.2898	0.2115	0.4599	66.15
4.8404	3.0	10410	4.8454	0.6901	4.1553	0.5401	0.3788	0.2917	0.5401	63.77
4.7439	4.0	13880	4.8177	0.6904	4.1274	0.5401	0.3788	0.2917	0.5401	62.02
4.6667	5.0	17350	4.8065	0.6903	4.1163	0.5401	0.3788	0.2917	0.5401	61.33
4.6018	6.0	20820	4.8081	0.6963	4.1119	0.5401	0.3788	0.2917	0.5401	61.06
4.5447	7.0	24290	4.8089	0.6912	4.1177	0.5401	0.3788	0.2917	0.5401	61.42
4.4944	8.0	27760	4.8128	0.6900	4.1228	0.5401	0.3788	0.2917	0.5401	61.73
4.4505	9.0	31230	4.8152	0.6905	4.1248	0.5401	0.3788	0.2917	0.5401	61.85
4.4116	10.0	34700	4.8129	0.6908	4.1221	0.5401	0.3788	0.2917	0.5401	61.69
4.3787	11.0	38170	4.8146	0.6906	4.1241	0.5401	0.3788	0.2917	0.5401	61.81
4.3494	12.0	41640	4.8229	0.6900	4.1329	0.5401	0.3788	0.2917	0.5401	62.36
4.3253	13.0	45110	4.8287	0.6900	4.1388	0.5401	0.3788	0.2917	0.5401	62.73
4.3075	14.0	48580	4.8247	0.6900	4.1347	0.5401	0.3788	0.2917	0.5401	62.47
4.2936	15.0	52050	4.8322	0.6900	4.1423	0.5401	0.3788	0.2917	0.5401	62.95

Framework versions

Transformers 4.21.2
Pytorch 1.12.1
Datasets 2.4.0
Tokenizers 0.12.1

Downloads last month: 5