Update README.md

README.md

@@ -10,27 +10,25 @@ datasets:
 
 UNDER CONSTRUCTION
 
-You can test this model at: https://huggingface.co/spaces/schuler/experimental-kphi-3-nano-4k-instruct-gradio-autoloader .
+You can test this model at: 
+* https://huggingface.co/spaces/schuler/experimental-kphi-3-nano-4k-instruct-gradio-autoloader .
+* https://huggingface.co/spaces/schuler/experimental-KPhi-3-nano-4k-instruct .
 
 ## Usage:
 ```
 from transformers import AutoTokenizer, AutoModelForCausalLM, GenerationConfig, pipeline
+from transformers import LlamaTokenizer
 import torch
 
 REPO_NAME = 'schuler/experimental-JP47D20'
 
-# How to cache?
-@spaces.GPU()
 def load_model(repo_name):
-    tokenizer = AutoTokenizer.from_pretrained(repo_name, trust_remote_code=True)
-    generator_conf = GenerationConfig.from_pretrained(repo_name)
-    model = AutoModelForCausalLM.from_pretrained(repo_name, trust_remote_code=True, torch_dtype=torch.bfloat16, attn_implementation="eager")
+    tokenizer = LlamaTokenizer.from_pretrained(REPO_NAME, trust_remote_code=True)
+    generator_conf = GenerationConfig.from_pretrained(REPO_NAME)
+    model = AutoModelForCausalLM.from_pretrained(REPO_NAME, trust_remote_code=True, torch_dtype=torch.bfloat16, attn_implementation="eager")
     # model.to('cuda')
     return tokenizer, generator_conf, model
 
-# tokenizer, generator_conf, model, generator = False, False, False, False 
-# with gr.Blocks() as main_block:
-
 tokenizer, generator_conf, model = load_model(REPO_NAME)
 global_error = ''
 try: