Azaz666
/

SmolVLM-Instruct-INT4-quantized

Model card Files Files and versions

SmolVLM-Instruct-INT4-quantized / generation_config.json

Azaz666's picture

INT4 quantization of HuggingFaceTB/SmolVLM-Instruct (LLM backbone quantized, vision encoder fp16)

113fe69 verified 2 months ago

history blame contribute delete

135 Bytes

	{
	"_from_model_config": true,
	"bos_token_id": 0,
	"eos_token_id": 49154,
	"pad_token_id": 2,
	"transformers_version": "5.3.0"
	}