| { | |
| "model_lib": "zephyr-7b-beta-chatRDM-q4f32_1", | |
| "local_id": "zephyr-7b-beta-q4f32_1", | |
| "conv_template": "mistral_default", | |
| "temperature": 0.7, | |
| "repetition_penalty": 1.0, | |
| "top_p": 0.95, | |
| "mean_gen_len": 128, | |
| "max_gen_len": 512, | |
| "num_shards": 1, | |
| "use_presharded_weights": false, | |
| "shift_fill_factor": 0.3, | |
| "conv_config": { | |
| "seps": [ | |
| " " | |
| ], | |
| "stop_tokens": [ | |
| 2 | |
| ], | |
| "offset": 0, | |
| "messages": [ | |
| ], | |
| "stop_str": "</s>", | |
| "roles": [ | |
| "<|user|>", | |
| "<|assistant|>" | |
| ], | |
| "role_msg_sep": " ", | |
| "role_empty_sep": " ", | |
| "system": "<|system|> A chat between a curious user and an artificial intelligence research data management assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. [/INST]", | |
| "add_bos": true, | |
| "name": "Llama2" | |
| }, | |
| "tokenizer_files": [ | |
| "tokenizer.json", | |
| "added_tokens.json", | |
| "tokenizer.model" | |
| ], | |
| "model_category": "mistral", | |
| "model_name": "zephyr-7b-beta", | |
| "vocab_size": 32000, | |
| "prefill_chunk_size": 4096, | |
| "sliding_window": 4096 | |
| } | |