Instructions to use Sarmistha/ganga-idiom-finetune with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Sarmistha/ganga-idiom-finetune with Transformers:

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

pipe = pipeline("text-generation", model="Sarmistha/ganga-idiom-finetune")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Sarmistha/ganga-idiom-finetune")
model = AutoModelForCausalLM.from_pretrained("Sarmistha/ganga-idiom-finetune", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Sarmistha/ganga-idiom-finetune with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/Sarmistha/ganga-idiom-finetune

SGLang

How to use Sarmistha/ganga-idiom-finetune 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 "Sarmistha/ganga-idiom-finetune" \
    --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": "Sarmistha/ganga-idiom-finetune",
		"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 "Sarmistha/ganga-idiom-finetune" \
        --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": "Sarmistha/ganga-idiom-finetune",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Sarmistha/ganga-idiom-finetune with Docker Model Runner:
```
docker model run hf.co/Sarmistha/ganga-idiom-finetune
```

ganga-idiom-finetune

This model is a fine-tuned version of LingoIITGN/ganga-1b on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 1.6450
Rouge1: 0.4650
Rouge2: 0.2624
Rougel: 0.4409
Rougelsum: 0.4427
Bleu1: 43.8834
Bleu2: 24.8997
Bleu3: 17.7287
Bleu: 22.3666
Moverscore: 0.0510
Cosine Distance: 0.4047
Jaccard Similarity: 0.2798
L1 Distance: 72.1289
L2 Distance: 8.9081
Freshman Readability: 12.9088

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: 5e-05
train_batch_size: 1
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 8
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 3
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Bleu1	Bleu2	Bleu3	Bleu	Moverscore	Cosine Distance	Jaccard Similarity	L1 Distance	L2 Distance	Freshman Readability
No log	1.0	112	1.6232	0.4393	0.2266	0.4064	0.4082	41.1287	21.6852	14.7151	19.1573	0.0221	0.4361	0.2462	78.5859	9.3605	11.1333
No log	2.0	224	1.6413	0.4519	0.2467	0.4322	0.4336	42.7056	23.3160	16.0230	20.6104	0.0356	0.4215	0.2666	75.7734	9.1753	12.5991
No log	2.9765	333	1.6590	0.4578	0.2522	0.4370	0.4382	43.3439	23.9543	16.4120	21.0878	0.0413	0.4147	0.2723	73.9766	9.0579	12.6140

Framework versions

Transformers 4.51.3
Pytorch 2.7.0+cu118
Datasets 3.6.0
Tokenizers 0.21.1

Downloads last month: 9

Safetensors

Model size

1.0B params

Tensor type

F32

Model tree for Sarmistha/ganga-idiom-finetune

Base model

LingoIITGN/ganga-1b

Finetuned

(3)

this model