Update app.py

app.py

@@ -6,15 +6,16 @@ import json
 from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig, GemmaTokenizer,  StoppingCriteria, StoppingCriteriaList, GenerationConfig
 import os
 
-sft_model = "somosnlp/RecetasDeLaAbuela_mistral-7b-instruct-v0.2-bnb-4bit"
-base_model_name = "unsloth/Mistral-7B-Instruct-v0.2"
+#sft_model = "somosnlp/RecetasDeLaAbuela_mistral-7b-instruct-v0.2-bnb-4bit"
+#base_model_name = "unsloth/Mistral-7B-Instruct-v0.2"
 #sft_model = "somosnlp/RecetasDeLaAbuela_gemma-2b-it-bnb-4bit"
 #base_model_name = "unsloth/gemma-2b-it-bnb-4bit"
 
-#sft_model = "somosnlp/RecetasDeLaAbuela5k_gemma-2b-bnb-4bit"
+sft_model = "somosnlp/RecetasDeLaAbuela5k_gemma-2b-bnb-4bit"
 #base_model_name = "unsloth/gemma-2b-bnb-4bit"
+base_model_name = "unsloth/gemma-2b-it-bnb-4bit
 
-max_seq_length=700
+max_seq_length=300
 base_model = AutoModelForCausalLM.from_pretrained(base_model_name,return_dict=True,device_map="auto", torch_dtype=torch.float16,)
 tokenizer = AutoTokenizer.from_pretrained(base_model_name, max_length = max_seq_length)
 ft_model = PeftModel.from_pretrained(base_model, sft_model)