Update README.md

README.md

@@ -11,8 +11,8 @@ language:
 
 Added a lot more data to sft, now json and multiturn work more stable on long context and hard prompts
 
- - [Google Colab](https://colab.research.google.com/drive/15O9LwZhVUa1LWhZa2UKr_B-KOKenJBvv#scrollTo=5EeNFU2-9ERi)
- - [GGUF](https://huggingface.co/Vikhrmodels/Vikhr-7B-instruct_0.4-GGUF)
+ - [Google Colab](https://colab.research.google.com/drive/1-_BWsJycBm3rEyjpBx2_ejshpemQYHbe?usp=sharing)
+ - [GGUF](https://huggingface.co/Vikhrmodels/it-5.2-fp16-cp-GGUF)
 
 ```python
 
@@ -24,7 +24,7 @@ model = AutoModelForCausalLM.from_pretrained("Vikhrmodels/it-5.2-fp16-cp",
                                              attn_implementation="flash_attention_2",
                                              torch_dtype=torch.bfloat16)
 
-tokenizer = AutoTokenizer.from_pretrained("Vikhrmodels/it-5.2-fp16-cp")
+tokenizer = AutoTokenizer.from_pretrained("Vikhrmodels/Vikhr-7B-instruct_0.4")
 from transformers import  AutoTokenizer, pipeline
 pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)
 prompts = [