Instructions to use cyttic/trocr-bigram5-lentrunc with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use cyttic/trocr-bigram5-lentrunc with Transformers:

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

pipe = pipeline("image-text-to-text", model="cyttic/trocr-bigram5-lentrunc")

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("cyttic/trocr-bigram5-lentrunc")
model = AutoModelForMultimodalLM.from_pretrained("cyttic/trocr-bigram5-lentrunc", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use cyttic/trocr-bigram5-lentrunc with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "cyttic/trocr-bigram5-lentrunc"
# 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-bigram5-lentrunc",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/cyttic/trocr-bigram5-lentrunc

SGLang

How to use cyttic/trocr-bigram5-lentrunc 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-bigram5-lentrunc" \
    --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-bigram5-lentrunc",
		"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-bigram5-lentrunc" \
        --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-bigram5-lentrunc",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use cyttic/trocr-bigram5-lentrunc with Docker Model Runner:
```
docker model run hf.co/cyttic/trocr-bigram5-lentrunc
```

trocr-bigram5-lentrunc

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: 0.3604
Cer: 0.0205
Wer: 0.0584

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
4.7708	0.1290	2000	2.3792	0.2069	0.4211
3.5688	0.2581	4000	1.6457	0.1326	0.3034
2.6281	0.3871	6000	1.2726	0.1018	0.2408
2.4209	0.5161	8000	1.0654	0.0818	0.1989
1.8799	0.6452	10000	0.9121	0.0681	0.1708
1.7538	0.7742	12000	0.8169	0.0571	0.1464
1.6138	0.9032	14000	0.7307	0.0527	0.1356
1.2246	1.0323	16000	0.6595	0.0443	0.1165
1.1681	1.1613	18000	0.6062	0.0403	0.1089
1.0557	1.2903	20000	0.5747	0.0378	0.1019
0.9707	1.4194	22000	0.5402	0.0343	0.0923
0.9284	1.5484	24000	0.5028	0.0316	0.0872
0.8654	1.6774	26000	0.4761	0.0302	0.0819
0.8459	1.8065	28000	0.4524	0.0271	0.0781
0.8298	1.9355	30000	0.4345	0.0256	0.0712
0.5770	2.0645	32000	0.4196	0.0262	0.0712
0.5775	2.1935	34000	0.4062	0.0263	0.0704
0.4802	2.3226	36000	0.3955	0.0242	0.0661
0.6221	2.4516	38000	0.3833	0.0228	0.0633
0.5440	2.5806	40000	0.3799	0.0224	0.0629
0.6317	2.7097	42000	0.3702	0.0214	0.0608
0.5272	2.8387	44000	0.3645	0.0212	0.0605
0.5501	2.9677	46000	0.3605	0.0204	0.0584
0.5624	3.0	46500	0.3604	0.0205	0.0584

Framework versions

Transformers 5.15.0
Pytorch 2.11.0+cu128
Datasets 5.0.1
Tokenizers 0.22.2

Downloads last month: 34

Safetensors

Model size

0.3B params

Tensor type

F32

Model tree for cyttic/trocr-bigram5-lentrunc

Base model

cyttic/exp2-frozen-benyehuda-cont

Finetuned

(51)

this model