Upload README.md with huggingface_hub

README.md

@@ -286,36 +286,6 @@ with torch.no_grad():
 print(tokenizer.decode(output_ids[0], skip_special_tokens=True))
 ```
 
-If the model files are stored at the repository root, remove `subfolder="hf"`:
-
-```python
-import torch
-from transformers import PreTrainedTokenizerFast, Llama4ForCausalLM
-
-repo = "shibatch/tinyllama4moe2m"
-
-tokenizer = PreTrainedTokenizerFast.from_pretrained(repo)
-model = Llama4ForCausalLM.from_pretrained(
-    repo,
-    torch_dtype=torch.float32,
-)
-model.eval()
-
-prompt = "Once upon"
-inputs = tokenizer(prompt, return_tensors="pt")
-
-with torch.no_grad():
-    output_ids = model.generate(
-        **inputs,
-        max_new_tokens=100,
-        do_sample=False,
-        pad_token_id=tokenizer.pad_token_id,
-        eos_token_id=tokenizer.eos_token_id,
-    )
-
-print(tokenizer.decode(output_ids[0], skip_special_tokens=True))
-```
-
 ## Loading requirements
 
 This checkpoint requires a Transformers version that supports Llama4 and Llama4 MoE.