temp

.gitignore

@@ -0,0 +1 @@
+./api.py

api.py

@@ -0,0 +1,8 @@
+from gradio_client import Client
+
+client = Client("as-cle-bert/bloom-multilingual-chat")
+result = client.predict(
+		message="Hello!!",
+		api_name="/chat"
+)
+print(result)

load_model.py

@@ -3,4 +3,4 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
 model = AutoModelForCausalLM.from_pretrained("bigscience/bloom-1b7")
 tokenizer = AutoTokenizer.from_pretrained("bigscience/bloom-1b7")
 
-pipe = pipeline("text-generation", model=model, tokenizer=tokenizer, max_new_tokens=2048, repetition_penalty=1.2)
+pipe = pipeline("text-generation", model=model, tokenizer=tokenizer, max_new_tokens=2048, repetition_penalty=1.2, temperature=0.4)