Instructions to use Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct with Transformers:

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

pipe = pipeline("image-text-to-text", model="Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct")
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 AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct")
model = AutoModelForCausalLM.from_pretrained("Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct", device_map="auto")
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?"}
        ]
    },
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct",
		"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/Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct

SGLang

How to use Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct 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 "Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct" \
    --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": "Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct",
		"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 "Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct" \
        --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": "Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct",
		"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 Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct with Docker Model Runner:
```
docker model run hf.co/Wojtekb30/Bielik-1.5B-v3.0-VLM-Instruct
```

Wojtekb30 commited on Mar 3

Commit

8350b14

1 Parent(s): 8c82385

Move model files to root directory

Browse files

Files changed (12) hide show

bielik_1.5b_vlm_merged/config.json +0 -35
bielik_1.5b_vlm_merged/chat_template.jinja → chat_template.jinja +3 -3
bielik_1.5b_vlm_merged/generation_config.json → generation_config.json +10 -10
bielik_1.5b_vlm_merged/mm_projector.safetensors → mm_projector.safetensors +0 -0
bielik_1.5b_vlm_merged/model.safetensors → model.safetensors +0 -0
model_index.json +0 -13
bielik_1.5b_vlm_merged/tokenizer.json → tokenizer.json +0 -0
bielik_1.5b_vlm_merged/tokenizer_config.json → tokenizer_config.json +22 -22
uruchom1.py +2 -2
uruchom1_GUIpicker.py +2 -2
uruchom1_fullGUI.py +2 -2
{bielik_1.5b_vlm_merged/vision_tower_config → vision_tower_config}/config.json +38 -38

bielik_1.5b_vlm_merged/config.json DELETED Viewed

@@ -1,35 +0,0 @@
-{
-  "architectures": [
-    "LlamaForCausalLM"
-  ],
-  "attention_bias": true,
-  "attention_dropout": 0.0,
-  "bos_token_id": 1,
-  "dtype": "float16",
-  "eos_token_id": [
-    4,
-    2
-  ],
-  "head_dim": 128,
-  "hidden_act": "silu",
-  "hidden_size": 1536,
-  "initializer_range": 0.02,
-  "intermediate_size": 8960,
-  "max_position_embeddings": 8192,
-  "mlp_bias": true,
-  "model_type": "llama",
-  "num_attention_heads": 12,
-  "num_hidden_layers": 32,
-  "num_key_value_heads": 2,
-  "pad_token_id": 2,
-  "pretraining_tp": 1,
-  "rms_norm_eps": 1e-06,
-  "rope_parameters": {
-    "rope_theta": 1000000,
-    "rope_type": "default"
-  },
-  "tie_word_embeddings": false,
-  "transformers_version": "5.1.0",
-  "use_cache": true,
-  "vocab_size": 32000
-}

bielik_1.5b_vlm_merged/chat_template.jinja → chat_template.jinja RENAMED Viewed

@@ -1,4 +1,4 @@
-{{bos_token}}{% for message in messages %}{{'<|im_start|>' + message['role'] + '
-' + message['content'] + '<|im_end|>' + '
-'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant
 ' }}{% endif %}

+{{bos_token}}{% for message in messages %}{{'<|im_start|>' + message['role'] + '
+' + message['content'] + '<|im_end|>' + '
+'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant
 ' }}{% endif %}

bielik_1.5b_vlm_merged/generation_config.json → generation_config.json RENAMED Viewed

@@ -1,10 +1,10 @@
-{
-  "_from_model_config": true,
-  "bos_token_id": 1,
-  "eos_token_id": [
-    4,
-    2
-  ],
-  "pad_token_id": 2,
-  "transformers_version": "5.1.0"
-}

+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": [
+    4,
+    2
+  ],
+  "pad_token_id": 2,
+  "transformers_version": "5.1.0"
+}

bielik_1.5b_vlm_merged/mm_projector.safetensors → mm_projector.safetensors RENAMED Viewed

File without changes

bielik_1.5b_vlm_merged/model.safetensors → model.safetensors RENAMED Viewed

File without changes

model_index.json DELETED Viewed

@@ -1,13 +0,0 @@
-{
-  "_class_name": "LlamaForCausalLM",
-  "_transformers_version": "5.1.0",
-  "model": [
-    "transformers",
-    "LlamaForCausalLM"
-  ],
-  "tokenizer": [
-    "transformers",
-    "LlamaTokenizerFast"
-  ],
-  "subfolder": "bielik_1.5b_vlm_merged"
-}

bielik_1.5b_vlm_merged/tokenizer.json → tokenizer.json RENAMED Viewed

File without changes

bielik_1.5b_vlm_merged/tokenizer_config.json → tokenizer_config.json RENAMED Viewed

@@ -1,22 +1,22 @@
-{
-  "add_prefix_space": null,
-  "backend": "tokenizers",
-  "bos_token": "<s>",
-  "clean_up_tokenization_spaces": false,
-  "eos_token": "<|im_end|>",
-  "extra_special_tokens": [
-    "<|im_start|>",
-    "<|im_end|>",
-    "<tool_call>",
-    "</tool_call>"
-  ],
-  "is_local": true,
-  "legacy": true,
-  "model_max_length": 1000000000000000019884624838656,
-  "pad_token": "</s>",
-  "sp_model_kwargs": {},
-  "spaces_between_special_tokens": false,
-  "tokenizer_class": "TokenizersBackend",
-  "unk_token": "<unk>",
-  "use_default_system_prompt": false
-}

+{
+  "add_prefix_space": null,
+  "backend": "tokenizers",
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<tool_call>",
+    "</tool_call>"
+  ],
+  "is_local": true,
+  "legacy": true,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "</s>",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

uruchom1.py CHANGED Viewed

@@ -13,10 +13,10 @@ VISION_MODEL_PATH = "google/siglip-so400m-patch14-384"
 #VISION_MODEL_PATH = "siglip-so400m-patch14-384"
 # Ścieżka do scalonego modelu językowego (LLM + adaptery)
-MERGED_MODEL_PATH = "./bielik_1.5b_vlm_merged"
 # Ścieżka do pliku z wagami projektora multimodalnego
-PROJECTOR_FILE = "./bielik_1.5b_vlm_merged/mm_projector.safetensors"
 # Automatyczny wybór urządzenia: GPU (CUDA) jeśli dostępne, w przeciwnym razie CPU
 DEVICE = "cuda" if torch.cuda.is_available() else "cpu"

 #VISION_MODEL_PATH = "siglip-so400m-patch14-384"
 # Ścieżka do scalonego modelu językowego (LLM + adaptery)
+MERGED_MODEL_PATH = "./"
 # Ścieżka do pliku z wagami projektora multimodalnego
+PROJECTOR_FILE = "./mm_projector.safetensors"
 # Automatyczny wybór urządzenia: GPU (CUDA) jeśli dostępne, w przeciwnym razie CPU
 DEVICE = "cuda" if torch.cuda.is_available() else "cpu"

uruchom1_GUIpicker.py CHANGED Viewed

@@ -15,10 +15,10 @@ VISION_MODEL_PATH = "google/siglip-so400m-patch14-384"
 #VISION_MODEL_PATH = "siglip-so400m-patch14-384"
 # Ścieżka do scalonego modelu językowego (LLM + adaptery)
-MERGED_MODEL_PATH = "./bielik_1.5b_vlm_merged"
 # Ścieżka do pliku z wagami projektora multimodalnego
-PROJECTOR_FILE = "./bielik_1.5b_vlm_merged/mm_projector.safetensors"
 # Automatyczny wybór urządzenia: GPU (CUDA) jeśli dostępne, w przeciwnym razie CPU
 DEVICE = "cuda" if torch.cuda.is_available() else "cpu"

 #VISION_MODEL_PATH = "siglip-so400m-patch14-384"
 # Ścieżka do scalonego modelu językowego (LLM + adaptery)
+MERGED_MODEL_PATH = "./"
 # Ścieżka do pliku z wagami projektora multimodalnego
+PROJECTOR_FILE = "./mm_projector.safetensors"
 # Automatyczny wybór urządzenia: GPU (CUDA) jeśli dostępne, w przeciwnym razie CPU
 DEVICE = "cuda" if torch.cuda.is_available() else "cpu"

uruchom1_fullGUI.py CHANGED Viewed

@@ -15,8 +15,8 @@ from safetensors.torch import load_file
 # Ścieżki / nazwy modeli
 VISION_MODEL_PATH = "google/siglip-so400m-patch14-384"
 #VISION_MODEL_PATH = "siglip-so400m-patch14-384"
-MERGED_MODEL_PATH = "./bielik_1.5b_vlm_merged"
-PROJECTOR_FILE = "./bielik_1.5b_vlm_merged/mm_projector.safetensors"
 DEVICE = "cuda" if torch.cuda.is_available() else "cpu"

 # Ścieżki / nazwy modeli
 VISION_MODEL_PATH = "google/siglip-so400m-patch14-384"
 #VISION_MODEL_PATH = "siglip-so400m-patch14-384"
+MERGED_MODEL_PATH = "./"
+PROJECTOR_FILE = "./mm_projector.safetensors"
 DEVICE = "cuda" if torch.cuda.is_available() else "cpu"

{bielik_1.5b_vlm_merged/vision_tower_config → vision_tower_config}/config.json RENAMED Viewed

@@ -1,38 +1,38 @@
-{
-  "architectures": [
-    "SiglipModel"
-  ],
-  "dtype": "float32",
-  "initializer_factor": 1.0,
-  "model_type": "siglip",
-  "text_config": {
-    "attention_dropout": 0.0,
-    "bos_token_id": 49406,
-    "eos_token_id": 49407,
-    "hidden_act": "gelu_pytorch_tanh",
-    "hidden_size": 1152,
-    "intermediate_size": 4304,
-    "layer_norm_eps": 1e-06,
-    "max_position_embeddings": 64,
-    "model_type": "siglip_text_model",
-    "num_attention_heads": 16,
-    "num_hidden_layers": 27,
-    "pad_token_id": 1,
-    "projection_size": 1152,
-    "vocab_size": 32000
-  },
-  "transformers_version": "5.1.0",
-  "vision_config": {
-    "attention_dropout": 0.0,
-    "hidden_act": "gelu_pytorch_tanh",
-    "hidden_size": 1152,
-    "image_size": 384,
-    "intermediate_size": 4304,
-    "layer_norm_eps": 1e-06,
-    "model_type": "siglip_vision_model",
-    "num_attention_heads": 16,
-    "num_channels": 3,
-    "num_hidden_layers": 27,
-    "patch_size": 14
-  }
-}

+{
+  "architectures": [
+    "SiglipModel"
+  ],
+  "dtype": "float32",
+  "initializer_factor": 1.0,
+  "model_type": "siglip",
+  "text_config": {
+    "attention_dropout": 0.0,
+    "bos_token_id": 49406,
+    "eos_token_id": 49407,
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 1152,
+    "intermediate_size": 4304,
+    "layer_norm_eps": 1e-06,
+    "max_position_embeddings": 64,
+    "model_type": "siglip_text_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 27,
+    "pad_token_id": 1,
+    "projection_size": 1152,
+    "vocab_size": 32000
+  },
+  "transformers_version": "5.1.0",
+  "vision_config": {
+    "attention_dropout": 0.0,
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 1152,
+    "image_size": 384,
+    "intermediate_size": 4304,
+    "layer_norm_eps": 1e-06,
+    "model_type": "siglip_vision_model",
+    "num_attention_heads": 16,
+    "num_channels": 3,
+    "num_hidden_layers": 27,
+    "patch_size": 14
+  }
+}