Image-Text-to-Text
Transformers
Safetensors
dots_ocr
text-generation
fp8
compressed-tensors
llm-compressor
quantized
multimodal
conversational
custom_code
Instructions to use binedge/dots.mocr-FP8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use binedge/dots.mocr-FP8 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="binedge/dots.mocr-FP8", trust_remote_code=True) messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] pipe(text=messages)# Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("binedge/dots.mocr-FP8", trust_remote_code=True, dtype="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use binedge/dots.mocr-FP8 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "binedge/dots.mocr-FP8" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "binedge/dots.mocr-FP8", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "Describe this image in one sentence." }, { "type": "image_url", "image_url": { "url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg" } } ] } ] }'Use Docker
docker model run hf.co/binedge/dots.mocr-FP8
- SGLang
How to use binedge/dots.mocr-FP8 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 "binedge/dots.mocr-FP8" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "binedge/dots.mocr-FP8", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "Describe this image in one sentence." }, { "type": "image_url", "image_url": { "url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg" } } ] } ] }'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 "binedge/dots.mocr-FP8" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "binedge/dots.mocr-FP8", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "Describe this image in one sentence." }, { "type": "image_url", "image_url": { "url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg" } } ] } ] }' - Docker Model Runner
How to use binedge/dots.mocr-FP8 with Docker Model Runner:
docker model run hf.co/binedge/dots.mocr-FP8
| { | |
| "architectures": [ | |
| "DotsOCRForCausalLM" | |
| ], | |
| "attention_bias": true, | |
| "attention_dropout": 0.0, | |
| "auto_map": { | |
| "AutoConfig": "configuration_dots.DotsOCRConfig", | |
| "AutoModelForCausalLM": "modeling_dots_ocr.DotsOCRForCausalLM" | |
| }, | |
| "dtype": "bfloat16", | |
| "hidden_act": "silu", | |
| "hidden_size": 1536, | |
| "image_token_id": 151665, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 8960, | |
| "layer_types": [ | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention" | |
| ], | |
| "max_position_embeddings": 131072, | |
| "max_window_layers": 28, | |
| "model_type": "dots_ocr", | |
| "num_attention_heads": 12, | |
| "num_hidden_layers": 28, | |
| "num_key_value_heads": 2, | |
| "quantization_config": { | |
| "config_groups": { | |
| "group_0": { | |
| "format": "float-quantized", | |
| "input_activations": { | |
| "actorder": null, | |
| "block_structure": null, | |
| "dynamic": true, | |
| "group_size": null, | |
| "num_bits": 8, | |
| "observer": null, | |
| "observer_kwargs": {}, | |
| "scale_dtype": null, | |
| "strategy": "token", | |
| "symmetric": true, | |
| "type": "float", | |
| "zp_dtype": null | |
| }, | |
| "output_activations": null, | |
| "targets": [ | |
| "Linear" | |
| ], | |
| "weights": { | |
| "actorder": null, | |
| "block_structure": null, | |
| "dynamic": false, | |
| "group_size": null, | |
| "num_bits": 8, | |
| "observer": "memoryless_minmax", | |
| "observer_kwargs": {}, | |
| "scale_dtype": null, | |
| "strategy": "channel", | |
| "symmetric": true, | |
| "type": "float", | |
| "zp_dtype": null | |
| } | |
| } | |
| }, | |
| "format": "float-quantized", | |
| "global_compression_ratio": null, | |
| "ignore": [ | |
| "lm_head", | |
| "vision_tower.blocks.0.attn.qkv", | |
| "vision_tower.blocks.0.attn.proj", | |
| "vision_tower.blocks.0.mlp.fc1", | |
| "vision_tower.blocks.0.mlp.fc2", | |
| "vision_tower.blocks.0.mlp.fc3", | |
| "vision_tower.blocks.1.attn.qkv", | |
| "vision_tower.blocks.1.attn.proj", | |
| "vision_tower.blocks.1.mlp.fc1", | |
| "vision_tower.blocks.1.mlp.fc2", | |
| "vision_tower.blocks.1.mlp.fc3", | |
| "vision_tower.blocks.2.attn.qkv", | |
| "vision_tower.blocks.2.attn.proj", | |
| "vision_tower.blocks.2.mlp.fc1", | |
| "vision_tower.blocks.2.mlp.fc2", | |
| "vision_tower.blocks.2.mlp.fc3", | |
| "vision_tower.blocks.3.attn.qkv", | |
| "vision_tower.blocks.3.attn.proj", | |
| "vision_tower.blocks.3.mlp.fc1", | |
| "vision_tower.blocks.3.mlp.fc2", | |
| "vision_tower.blocks.3.mlp.fc3", | |
| "vision_tower.blocks.4.attn.qkv", | |
| "vision_tower.blocks.4.attn.proj", | |
| "vision_tower.blocks.4.mlp.fc1", | |
| "vision_tower.blocks.4.mlp.fc2", | |
| "vision_tower.blocks.4.mlp.fc3", | |
| "vision_tower.blocks.5.attn.qkv", | |
| "vision_tower.blocks.5.attn.proj", | |
| "vision_tower.blocks.5.mlp.fc1", | |
| "vision_tower.blocks.5.mlp.fc2", | |
| "vision_tower.blocks.5.mlp.fc3", | |
| "vision_tower.blocks.6.attn.qkv", | |
| "vision_tower.blocks.6.attn.proj", | |
| "vision_tower.blocks.6.mlp.fc1", | |
| "vision_tower.blocks.6.mlp.fc2", | |
| "vision_tower.blocks.6.mlp.fc3", | |
| "vision_tower.blocks.7.attn.qkv", | |
| "vision_tower.blocks.7.attn.proj", | |
| "vision_tower.blocks.7.mlp.fc1", | |
| "vision_tower.blocks.7.mlp.fc2", | |
| "vision_tower.blocks.7.mlp.fc3", | |
| "vision_tower.blocks.8.attn.qkv", | |
| "vision_tower.blocks.8.attn.proj", | |
| "vision_tower.blocks.8.mlp.fc1", | |
| "vision_tower.blocks.8.mlp.fc2", | |
| "vision_tower.blocks.8.mlp.fc3", | |
| "vision_tower.blocks.9.attn.qkv", | |
| "vision_tower.blocks.9.attn.proj", | |
| "vision_tower.blocks.9.mlp.fc1", | |
| "vision_tower.blocks.9.mlp.fc2", | |
| "vision_tower.blocks.9.mlp.fc3", | |
| "vision_tower.blocks.10.attn.qkv", | |
| "vision_tower.blocks.10.attn.proj", | |
| "vision_tower.blocks.10.mlp.fc1", | |
| "vision_tower.blocks.10.mlp.fc2", | |
| "vision_tower.blocks.10.mlp.fc3", | |
| "vision_tower.blocks.11.attn.qkv", | |
| "vision_tower.blocks.11.attn.proj", | |
| "vision_tower.blocks.11.mlp.fc1", | |
| "vision_tower.blocks.11.mlp.fc2", | |
| "vision_tower.blocks.11.mlp.fc3", | |
| "vision_tower.blocks.12.attn.qkv", | |
| "vision_tower.blocks.12.attn.proj", | |
| "vision_tower.blocks.12.mlp.fc1", | |
| "vision_tower.blocks.12.mlp.fc2", | |
| "vision_tower.blocks.12.mlp.fc3", | |
| "vision_tower.blocks.13.attn.qkv", | |
| "vision_tower.blocks.13.attn.proj", | |
| "vision_tower.blocks.13.mlp.fc1", | |
| "vision_tower.blocks.13.mlp.fc2", | |
| "vision_tower.blocks.13.mlp.fc3", | |
| "vision_tower.blocks.14.attn.qkv", | |
| "vision_tower.blocks.14.attn.proj", | |
| "vision_tower.blocks.14.mlp.fc1", | |
| "vision_tower.blocks.14.mlp.fc2", | |
| "vision_tower.blocks.14.mlp.fc3", | |
| "vision_tower.blocks.15.attn.qkv", | |
| "vision_tower.blocks.15.attn.proj", | |
| "vision_tower.blocks.15.mlp.fc1", | |
| "vision_tower.blocks.15.mlp.fc2", | |
| "vision_tower.blocks.15.mlp.fc3", | |
| "vision_tower.blocks.16.attn.qkv", | |
| "vision_tower.blocks.16.attn.proj", | |
| "vision_tower.blocks.16.mlp.fc1", | |
| "vision_tower.blocks.16.mlp.fc2", | |
| "vision_tower.blocks.16.mlp.fc3", | |
| "vision_tower.blocks.17.attn.qkv", | |
| "vision_tower.blocks.17.attn.proj", | |
| "vision_tower.blocks.17.mlp.fc1", | |
| "vision_tower.blocks.17.mlp.fc2", | |
| "vision_tower.blocks.17.mlp.fc3", | |
| "vision_tower.blocks.18.attn.qkv", | |
| "vision_tower.blocks.18.attn.proj", | |
| "vision_tower.blocks.18.mlp.fc1", | |
| "vision_tower.blocks.18.mlp.fc2", | |
| "vision_tower.blocks.18.mlp.fc3", | |
| "vision_tower.blocks.19.attn.qkv", | |
| "vision_tower.blocks.19.attn.proj", | |
| "vision_tower.blocks.19.mlp.fc1", | |
| "vision_tower.blocks.19.mlp.fc2", | |
| "vision_tower.blocks.19.mlp.fc3", | |
| "vision_tower.blocks.20.attn.qkv", | |
| "vision_tower.blocks.20.attn.proj", | |
| "vision_tower.blocks.20.mlp.fc1", | |
| "vision_tower.blocks.20.mlp.fc2", | |
| "vision_tower.blocks.20.mlp.fc3", | |
| "vision_tower.blocks.21.attn.qkv", | |
| "vision_tower.blocks.21.attn.proj", | |
| "vision_tower.blocks.21.mlp.fc1", | |
| "vision_tower.blocks.21.mlp.fc2", | |
| "vision_tower.blocks.21.mlp.fc3", | |
| "vision_tower.blocks.22.attn.qkv", | |
| "vision_tower.blocks.22.attn.proj", | |
| "vision_tower.blocks.22.mlp.fc1", | |
| "vision_tower.blocks.22.mlp.fc2", | |
| "vision_tower.blocks.22.mlp.fc3", | |
| "vision_tower.blocks.23.attn.qkv", | |
| "vision_tower.blocks.23.attn.proj", | |
| "vision_tower.blocks.23.mlp.fc1", | |
| "vision_tower.blocks.23.mlp.fc2", | |
| "vision_tower.blocks.23.mlp.fc3", | |
| "vision_tower.blocks.24.attn.qkv", | |
| "vision_tower.blocks.24.attn.proj", | |
| "vision_tower.blocks.24.mlp.fc1", | |
| "vision_tower.blocks.24.mlp.fc2", | |
| "vision_tower.blocks.24.mlp.fc3", | |
| "vision_tower.blocks.25.attn.qkv", | |
| "vision_tower.blocks.25.attn.proj", | |
| "vision_tower.blocks.25.mlp.fc1", | |
| "vision_tower.blocks.25.mlp.fc2", | |
| "vision_tower.blocks.25.mlp.fc3", | |
| "vision_tower.blocks.26.attn.qkv", | |
| "vision_tower.blocks.26.attn.proj", | |
| "vision_tower.blocks.26.mlp.fc1", | |
| "vision_tower.blocks.26.mlp.fc2", | |
| "vision_tower.blocks.26.mlp.fc3", | |
| "vision_tower.blocks.27.attn.qkv", | |
| "vision_tower.blocks.27.attn.proj", | |
| "vision_tower.blocks.27.mlp.fc1", | |
| "vision_tower.blocks.27.mlp.fc2", | |
| "vision_tower.blocks.27.mlp.fc3", | |
| "vision_tower.blocks.28.attn.qkv", | |
| "vision_tower.blocks.28.attn.proj", | |
| "vision_tower.blocks.28.mlp.fc1", | |
| "vision_tower.blocks.28.mlp.fc2", | |
| "vision_tower.blocks.28.mlp.fc3", | |
| "vision_tower.blocks.29.attn.qkv", | |
| "vision_tower.blocks.29.attn.proj", | |
| "vision_tower.blocks.29.mlp.fc1", | |
| "vision_tower.blocks.29.mlp.fc2", | |
| "vision_tower.blocks.29.mlp.fc3", | |
| "vision_tower.blocks.30.attn.qkv", | |
| "vision_tower.blocks.30.attn.proj", | |
| "vision_tower.blocks.30.mlp.fc1", | |
| "vision_tower.blocks.30.mlp.fc2", | |
| "vision_tower.blocks.30.mlp.fc3", | |
| "vision_tower.blocks.31.attn.qkv", | |
| "vision_tower.blocks.31.attn.proj", | |
| "vision_tower.blocks.31.mlp.fc1", | |
| "vision_tower.blocks.31.mlp.fc2", | |
| "vision_tower.blocks.31.mlp.fc3", | |
| "vision_tower.blocks.32.attn.qkv", | |
| "vision_tower.blocks.32.attn.proj", | |
| "vision_tower.blocks.32.mlp.fc1", | |
| "vision_tower.blocks.32.mlp.fc2", | |
| "vision_tower.blocks.32.mlp.fc3", | |
| "vision_tower.blocks.33.attn.qkv", | |
| "vision_tower.blocks.33.attn.proj", | |
| "vision_tower.blocks.33.mlp.fc1", | |
| "vision_tower.blocks.33.mlp.fc2", | |
| "vision_tower.blocks.33.mlp.fc3", | |
| "vision_tower.blocks.34.attn.qkv", | |
| "vision_tower.blocks.34.attn.proj", | |
| "vision_tower.blocks.34.mlp.fc1", | |
| "vision_tower.blocks.34.mlp.fc2", | |
| "vision_tower.blocks.34.mlp.fc3", | |
| "vision_tower.blocks.35.attn.qkv", | |
| "vision_tower.blocks.35.attn.proj", | |
| "vision_tower.blocks.35.mlp.fc1", | |
| "vision_tower.blocks.35.mlp.fc2", | |
| "vision_tower.blocks.35.mlp.fc3", | |
| "vision_tower.blocks.36.attn.qkv", | |
| "vision_tower.blocks.36.attn.proj", | |
| "vision_tower.blocks.36.mlp.fc1", | |
| "vision_tower.blocks.36.mlp.fc2", | |
| "vision_tower.blocks.36.mlp.fc3", | |
| "vision_tower.blocks.37.attn.qkv", | |
| "vision_tower.blocks.37.attn.proj", | |
| "vision_tower.blocks.37.mlp.fc1", | |
| "vision_tower.blocks.37.mlp.fc2", | |
| "vision_tower.blocks.37.mlp.fc3", | |
| "vision_tower.blocks.38.attn.qkv", | |
| "vision_tower.blocks.38.attn.proj", | |
| "vision_tower.blocks.38.mlp.fc1", | |
| "vision_tower.blocks.38.mlp.fc2", | |
| "vision_tower.blocks.38.mlp.fc3", | |
| "vision_tower.blocks.39.attn.qkv", | |
| "vision_tower.blocks.39.attn.proj", | |
| "vision_tower.blocks.39.mlp.fc1", | |
| "vision_tower.blocks.39.mlp.fc2", | |
| "vision_tower.blocks.39.mlp.fc3", | |
| "vision_tower.blocks.40.attn.qkv", | |
| "vision_tower.blocks.40.attn.proj", | |
| "vision_tower.blocks.40.mlp.fc1", | |
| "vision_tower.blocks.40.mlp.fc2", | |
| "vision_tower.blocks.40.mlp.fc3", | |
| "vision_tower.blocks.41.attn.qkv", | |
| "vision_tower.blocks.41.attn.proj", | |
| "vision_tower.blocks.41.mlp.fc1", | |
| "vision_tower.blocks.41.mlp.fc2", | |
| "vision_tower.blocks.41.mlp.fc3", | |
| "vision_tower.merger.mlp.0", | |
| "vision_tower.merger.mlp.2" | |
| ], | |
| "kv_cache_scheme": null, | |
| "quant_method": "compressed-tensors", | |
| "quantization_status": "compressed", | |
| "sparsity_config": {}, | |
| "transform_config": {}, | |
| "version": "0.15.1.a20260414" | |
| }, | |
| "rms_norm_eps": 1e-06, | |
| "rope_scaling": null, | |
| "rope_theta": 1000000, | |
| "sliding_window": null, | |
| "tie_word_embeddings": false, | |
| "transformers_version": "4.57.6", | |
| "use_cache": true, | |
| "use_sliding_window": false, | |
| "video_token_id": 151656, | |
| "vision_config": { | |
| "attn_implementation": "sdpa", | |
| "embed_dim": 1536, | |
| "gradient_checkpointing": false, | |
| "hidden_size": 1536, | |
| "init_merger_std": 0.02, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 4224, | |
| "is_causal": false, | |
| "model_type": "dots_vit", | |
| "num_attention_heads": 12, | |
| "num_channels": 3, | |
| "num_hidden_layers": 42, | |
| "patch_size": 14, | |
| "post_norm": true, | |
| "rms_norm_eps": 1e-05, | |
| "spatial_merge_size": 2, | |
| "temporal_patch_size": 1, | |
| "use_bias": false | |
| }, | |
| "vocab_size": 151936 | |
| } |