1e0d4f3753
* fix: correct phi3 * redundant phi2 dolphin * add: hermes llama3 * add: ngl settings * correct ctx len * correct ngl * correct maxlen + ngl * disable phi3 * add ngl * add ngl * add ngl * add ngl * add ngl * add ngl * add ngl * remove redundant hermes pro * add ngl * add ngl * add ngl * remove miqu * add ngl * add ngl * add ngl * add ngl * remove redundant * add ngl * add ngl * add ngl * add ngl * add ngl * add ngl * add ngl * add ngl * add ngl * version package bump * feat: resolve issue of cannot found model in the extensions due to the removal * feat: completely remove hermes-pro-7b * feat: completely remove openhermes-neural-7b and miqu-70b, and add llama3-hermes-8b via renaming from Rex * fix: correct description --------- Co-authored-by: Van-QA <van@jan.ai>
36 lines
921 B
JSON
36 lines
921 B
JSON
{
|
|
"sources": [
|
|
{
|
|
"filename": "llama-2-70b-chat.Q4_K_M.gguf",
|
|
"url": "https://huggingface.co/TheBloke/Llama-2-70B-Chat-GGUF/resolve/main/llama-2-70b-chat.Q4_K_M.gguf"
|
|
}
|
|
],
|
|
"id": "llama2-chat-70b",
|
|
"object": "model",
|
|
"name": "Llama 2 Chat 70B Q4",
|
|
"version": "1.0",
|
|
"description": "Llama 2 specifically designed for a comprehensive understanding the world.",
|
|
"format": "gguf",
|
|
"settings": {
|
|
"ctx_len": 4096,
|
|
"prompt_template": "[INST] <<SYS>>\n{system_message}<</SYS>>\n{prompt}[/INST]",
|
|
"llama_model_path": "llama-2-70b-chat.Q4_K_M.gguf",
|
|
"ngl": 80
|
|
},
|
|
"parameters": {
|
|
"temperature": 0.7,
|
|
"top_p": 0.95,
|
|
"stream": true,
|
|
"max_tokens": 4096,
|
|
"stop": [],
|
|
"frequency_penalty": 0,
|
|
"presence_penalty": 0
|
|
},
|
|
"metadata": {
|
|
"author": "MetaAI",
|
|
"tags": ["70B", "Foundational Model"],
|
|
"size": 43920000000
|
|
},
|
|
"engine": "nitro"
|
|
}
|