where are models here? Thanks.
processor = AutoProcessor.from_pretrained("./model/Qwen2.5-Thinker-3B-added-action-tokens")
processor.tokenizer.padding_side = 'left'
model = Qwen2_5OmniThinkerForConditionalGeneration.from_pretrained(
"./model/Qwen2.5-Thinker-3B-added-action-tokens",
torch_dtype=torch.bfloat16,
attn_implementation="flash_attention_2"
)
fast_tokenizer = AutoProcessor.from_pretrained(
"./model/pi_fast", trust_remote_code=True
)
where are models here? Thanks.