Instructions to use cyttic/trocr-bigram4-BY with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use cyttic/trocr-bigram4-BY with Transformers:

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

pipe = pipeline("image-text-to-text", model="cyttic/trocr-bigram4-BY")

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("cyttic/trocr-bigram4-BY")
model = AutoModelForMultimodalLM.from_pretrained("cyttic/trocr-bigram4-BY", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use cyttic/trocr-bigram4-BY with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "cyttic/trocr-bigram4-BY"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "cyttic/trocr-bigram4-BY",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/cyttic/trocr-bigram4-BY

SGLang

How to use cyttic/trocr-bigram4-BY 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 "cyttic/trocr-bigram4-BY" \
    --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": "cyttic/trocr-bigram4-BY",
		"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 "cyttic/trocr-bigram4-BY" \
        --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": "cyttic/trocr-bigram4-BY",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use cyttic/trocr-bigram4-BY with Docker Model Runner:
```
docker model run hf.co/cyttic/trocr-bigram4-BY
```

trocr-bigram4-BY

This model is a fine-tuned version of cyttic/exp2-frozen-benyehuda-cont on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.5704
Cer: 0.0297
Wer: 0.0866

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: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 16
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 4650
num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Cer	Wer
3.6881	0.1290	2000	1.6167	0.1867	0.3828
3.3020	0.2581	4000	1.3983	0.1329	0.2981
3.0724	0.3871	6000	1.2319	0.1036	0.2453
2.6093	0.5161	8000	1.1277	0.0854	0.2149
2.5148	0.6452	10000	1.0344	0.0731	0.1916
2.0941	0.7742	12000	0.9543	0.0600	0.1664
1.9738	0.9032	14000	0.8868	0.0577	0.1567
1.6005	1.0323	16000	0.8790	0.0553	0.1456
1.5485	1.1613	18000	0.8043	0.0488	0.1361
1.5126	1.2903	20000	0.7839	0.0460	0.1272
1.3922	1.4194	22000	0.7411	0.0443	0.1216
1.3501	1.5484	24000	0.7182	0.0395	0.1095
1.3266	1.6774	26000	0.6833	0.0380	0.1082
1.3512	1.8065	28000	0.6648	0.0354	0.0988
1.2295	1.9355	30000	0.6381	0.0344	0.0969
1.0388	2.0645	32000	0.6283	0.0352	0.0992
1.0065	2.1935	34000	0.6204	0.0360	0.0995
1.0404	2.3226	36000	0.6056	0.0329	0.0934
0.9850	2.4516	38000	0.5936	0.0318	0.0915
0.9691	2.5806	40000	0.5842	0.0305	0.0874
1.0368	2.7097	42000	0.5788	0.0298	0.0868
0.9213	2.8387	44000	0.5724	0.0303	0.0868
1.0162	2.9677	46000	0.5709	0.0302	0.0872
0.8843	3.0	46500	0.5704	0.0297	0.0866

Framework versions

Transformers 5.15.0
Pytorch 2.11.0+cu128
Datasets 5.0.1
Tokenizers 0.22.2

Downloads last month: 26

Safetensors

Model size

0.3B params

Tensor type

F32

Inference Providers NEW

Image-Text-to-Text

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for cyttic/trocr-bigram4-BY

Base model

cyttic/exp2-frozen-benyehuda-cont

Finetuned

(25)

this model