84a09ae03f
* fix(mistral-ins): clean redundant parameters * add(yarn-mistral): update new requested model * fix(trinity-v1): delete trinity v1 from the hub * add(tulu-2-70b): llama 70b alternative * fix(lzlv-70b): delete lzlv-70b and changed to tulu-2 * fix(mistral-ins): upgrade model version to v0.2 * fix(model-extention): pump version to 1.0.18 * add(dolphin 8x7b): update the current best moe finetuned model * add(openchat): the best 7b model * fix(tinyllama): pump version of the model to v1 * fix(stealth): upgrade stealth to v1.3 * Revert "fix(stealth): upgrade stealth to v1.3" This reverts commit da24df3fb5d69f93d92cc4dd45f991d548aff6aa. * fix(stealth): upgrade version to v1.3
22 lines
983 B
JSON
22 lines
983 B
JSON
{
|
|
"source_url": "https://huggingface.co/TheBloke/TinyLlama-1.1B-Chat-v1.0-GGUF/resolve/main/tinyllama-1.1b-chat-v1.0.Q4_K_M.gguf",
|
|
"id": "tinyllama-1.1b",
|
|
"object": "model",
|
|
"name": "TinyLlama Chat 1.1B Q4",
|
|
"version": "1.0",
|
|
"description": "The TinyLlama project, featuring a 1.1B parameter Llama model, is pretrained on an expansive 3 trillion token dataset. Its design ensures easy integration with various Llama-based open-source projects. Despite its smaller size, it efficiently utilizes lower computational and memory resources, drawing on GPT-4's analytical prowess to enhance its conversational abilities and versatility.",
|
|
"format": "gguf",
|
|
"settings": {
|
|
"ctx_len": 2048,
|
|
"prompt_template": "<|system|>\n{system_message}<|user|>\n{prompt}<|assistant|>"
|
|
},
|
|
"parameters": {
|
|
"max_tokens": 2048
|
|
},
|
|
"metadata": {
|
|
"author": "TinyLlama",
|
|
"tags": ["Tiny", "Foundation Model"],
|
|
"size": 669000000
|
|
},
|
|
"engine": "nitro"
|
|
} |