revert

app.py

@@ -10,10 +10,11 @@ from torchvision.transforms.v2 import Resize
 import subprocess
 subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
 
-model_id = "markury/motd"
-tokenizer = AutoTokenizer.from_pretrained(model_id)
+model_id = "vikhyatk/moondream2"
+revision = "2024-05-20"
+tokenizer = AutoTokenizer.from_pretrained(model_id, revision=revision)
 moondream = AutoModelForCausalLM.from_pretrained(
-    model_id, trust_remote_code=True,
+    model_id, trust_remote_code=True, revision=revision,
     torch_dtype=torch.bfloat16, device_map={"": "cuda"},
     attn_implementation="flash_attention_2"
 )