import os def init(cfg): print(cfg['setting_cache_path'].value) if os.path.exists(cfg['setting_cache_path'].value): # ========== 加载角色卡-缓存 ========== tmp = cfg['model'].load_session(cfg['setting_cache_path'].value) print(f"load cache from {cfg['setting_cache_path'].value} {tmp}") tmp = cfg['chat_template']('system', cfg['text_format'](cfg['role_char_d'].value, char=cfg['role_char'].value, user=cfg['role_usr'].value)) cfg['setting_n_keep'].value = len(tmp) tmp = cfg['chat_template'](cfg['role_char'].value, cfg['text_format'](cfg['role_chat_style'].value, char=cfg['role_char'].value, user=cfg['role_usr'].value)) cfg['setting_n_keep'].value += len(tmp) # ========== 加载角色卡-第一条消息 ========== cfg['chatbot'] = [] for one in cfg["role_char_first"]: one['name'] = cfg['text_format'](one['name'], char=cfg['role_char'].value, user=cfg['role_usr'].value) one['value'] = cfg['text_format'](one['value'], char=cfg['role_char'].value, user=cfg['role_usr'].value) if one['name'] == cfg['role_char'].value: cfg['chatbot'].append((None, cfg['chat_display_format'](one['value']))) print(one) else: # ========== 加载角色卡-角色描述 ========== tmp = cfg['chat_template']('system', cfg['text_format'](cfg['role_char_d'].value, char=cfg['role_char'].value, user=cfg['role_usr'].value)) cfg['setting_n_keep'].value = cfg['model'].eval_t(tmp) # 此内容永久存在 # ========== 加载角色卡-回复示例 ========== tmp = cfg['chat_template'](cfg['role_char'].value, cfg['text_format'](cfg['role_chat_style'].value, char=cfg['role_char'].value, user=cfg['role_usr'].value)) cfg['setting_n_keep'].value = cfg['model'].eval_t(tmp) # 此内容永久存在 # ========== 加载角色卡-第一条消息 ========== cfg['chatbot'] = [] for one in cfg["role_char_first"]: one['name'] = cfg['text_format'](one['name'], char=cfg['role_char'].value, user=cfg['role_usr'].value) one['value'] = cfg['text_format'](one['value'], char=cfg['role_char'].value, user=cfg['role_usr'].value) if one['name'] == cfg['role_char'].value: cfg['chatbot'].append((None, cfg['chat_display_format'](one['value']))) print(one) tmp = cfg['chat_template'](one['name'], one['value']) cfg['model'].eval_t(tmp) # 此内容随上下文增加将被丢弃 # ========== 保存角色卡-缓存 ========== with open(cfg['setting_cache_path'].value, 'wb') as f: pass tmp = cfg['model'].save_session(cfg['setting_cache_path'].value) print(f'save cache {tmp}') # ========== 上传缓存 ========== if os.environ.get("HF_TOKEN"): from huggingface_hub import login, CommitScheduler login(token=os.environ.get("HF_TOKEN"), write_permission=True) CommitScheduler(repo_id='Limour/llama-python-streamingllm-cache', repo_type='dataset', folder_path='cache')