Skip to content Skip to sidebar Skip to footer

octo-small

Edit model card Octo Small Octo Small See https://github.com/octo-models/octo for instructions for using this model. Octo Small is trained with a window size of 2, predicting 7-dimensional actions…

ARIA-7b-v0.1

Edit model card Training procedure The following bitsandbytes quantization config was used during training: quant_method: bitsandbytes load_in_8bit: False load_in_4bit: True llm_int8_threshold: 6.0 llm_int8_skip_modules: None llm_int8_enable_fp32_cpu_offload: False llm_int8_has_fp16_weight: False bnb_4bit_quant_type:…