Upload 4 files

README.md

@@ -1,13 +1 @@
----
-title: Ultimate 96B Chat
-emoji: 🐢
-colorFrom: yellow
-colorTo: yellow
-sdk: gradio
-sdk_version: 4.12.0
-app_file: app.py
-pinned: false
-license: mit
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# chat-hugginface

chat_bot.py

@@ -0,0 +1,33 @@
+import os
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
+# os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
+
+# モデルとトークナイザーのロード
+model_name = "cyberagent/open-calm-large"
+# model_name = "cyberagent/open-calm-3b"
+model = AutoModelForCausalLM.from_pretrained(model_name).to(torch.device("cuda" if torch.cuda.is_available() else "cpu"))
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+
+# チャットの開始
+print("ボット: こんにちは! 何か質問がありますか?")
+
+while True:
+    # ユーザからの入力を受け取る
+    user_input = input("あなた: ")
+
+    # ユーザの入力をエンコードしてテンソルに変換
+    input_ids = tokenizer.encode(user_input, return_tensors='pt')
+    attention_mask = (input_ids != tokenizer.pad_token_id if tokenizer.pad_token_id is not None else 0).int()
+
+    # 入力データを適切なデバイスに送る
+    input_ids = input_ids.to(model.device)
+    attention_mask = attention_mask.to(model.device)
+
+    # モデルによる応答の生成
+    output = model.generate(input_ids, attention_mask=attention_mask, max_length=300, num_return_sequences=1, no_repeat_ngram_size=2, pad_token_id=model.config.eos_token_id)
+
+    # 生成されたテキストのデコード
+    output_text = tokenizer.decode(output[:, input_ids.shape[-1]:][0], skip_special_tokens=True)
+
+    print("ボット: " + output_text)

chat_bot_v2.py

@@ -0,0 +1,12 @@
+# from cybermodel.model import set_model
+# from cybermodel.chatbot import run_bot
+from cybermodel import set_model, run_bot
+
+
+if __name__ == "__main__":
+    model_name = "cyberagent/open-calm-large"
+    model, tokenizer = set_model(model_name=model_name)
+
+    first_text = "ボット: ハロー！なんだい？"
+    you_text = "あなた："
+    run_bot(model, tokenizer, first_text, you_text, max_count=1)

chat_generate.py

@@ -0,0 +1,23 @@
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+# モデルとトークナイザーのロード
+model_name = "cyberagent/open-calm-large"
+# model_name = "cyberagent/open-calm-3b"
+model = AutoModelForCausalLM.from_pretrained(model_name).to(torch.device("cuda" if torch.cuda.is_available() else "cpu"))
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+
+inputs = tokenizer("Clinical AIは、", return_tensors="pt").to(model.device)
+with torch.no_grad():
+    tokens = model.generate(
+        **inputs,
+        max_new_tokens=64,
+        do_sample=True,
+        temperature=0.7,
+        top_p=0.9,
+        repetition_penalty=1.05,
+        pad_token_id=tokenizer.pad_token_id,
+    )
+    
+output = tokenizer.decode(tokens[0], skip_special_tokens=True)
+print(output)