Downloads · 30 days
95
4% of all-time downloads
RichardErkhov/Azure99_-_blossom-v4-qwen1_5-7b-gguf
Azure99_-_blossom-v4-qwen1_5-7b-gguf is a machine learning model from RichardErkhov. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Downloads · 30 days
95
4% of all-time downloads
All-time downloads
2.3K
Public
Repo size
89.9 GB
Likes
0
Public
Click a slice to open those files.
.gguf104 GB · 100%
From the Hugging Face model README
Quantization made by Richard Erkhov.
blossom-v4-qwen1_5-7b - GGUF
| Name | Quant method | Size |
|---|---|---|
| blossom-v4-qwen1_5-7b.Q2_K.gguf | Q2_K | 2.89GB |
| blossom-v4-qwen1_5-7b.IQ3_XS.gguf | IQ3_XS | 3.18GB |
| blossom-v4-qwen1_5-7b.IQ3_S.gguf | IQ3_S | 3.32GB |
| blossom-v4-qwen1_5-7b.Q3_K_S.gguf | Q3_K_S | 3.32GB |
| blossom-v4-qwen1_5-7b.IQ3_M.gguf | IQ3_M | 3.48GB |
| blossom-v4-qwen1_5-7b.Q3_K.gguf | Q3_K | 3.65GB |
| blossom-v4-qwen1_5-7b.Q3_K_M.gguf | Q3_K_M | 3.65GB |
| blossom-v4-qwen1_5-7b.Q3_K_L.gguf | Q3_K_L | 3.93GB |
| blossom-v4-qwen1_5-7b.IQ4_XS.gguf | IQ4_XS | 4.02GB |
| blossom-v4-qwen1_5-7b.Q4_0.gguf | Q4_0 | 4.2GB |
| blossom-v4-qwen1_5-7b.IQ4_NL.gguf | IQ4_NL | 4.22GB |
| blossom-v4-qwen1_5-7b.Q4_K_S.gguf | Q4_K_S | 4.23GB |
| blossom-v4-qwen1_5-7b.Q4_K.gguf | Q4_K | 4.44GB |
| blossom-v4-qwen1_5-7b.Q4_K_M.gguf | Q4_K_M | 4.44GB |
| blossom-v4-qwen1_5-7b.Q4_1.gguf | Q4_1 | 4.62GB |
| blossom-v4-qwen1_5-7b.Q5_0.gguf | Q5_0 | 5.03GB |
| blossom-v4-qwen1_5-7b.Q5_K_S.gguf | Q5_K_S | 5.03GB |
| blossom-v4-qwen1_5-7b.Q5_K.gguf | Q5_K | 5.15GB |
| blossom-v4-qwen1_5-7b.Q5_K_M.gguf | Q5_K_M | 5.15GB |
| blossom-v4-qwen1_5-7b.Q5_1.gguf | Q5_1 | 5.44GB |
| blossom-v4-qwen1_5-7b.Q6_K.gguf | Q6_K | 5.91GB |
| blossom-v4-qwen1_5-7b.Q8_0.gguf | Q8_0 | 7.65GB |
license: apache-2.0 datasets:
Blossom是一个对话式语言模型,基于Qwen1.5-7B预训练模型,在Blossom Orca/Wizard/Chat/Math混合数据集上进行指令精调得来。Blossom拥有强大的通用能力及上下文理解能力,此外,训练使用的高质量中英文数据集也进行了开源。
训练分为两阶段,第一阶段使用100K Wizard、100K Orca、20K Math单轮指令数据集,训练1个epoch;第二阶段使用50K Blossom chat多轮对话数据集、以及上一阶段中随机采样2%的数据,训练3个epoch。
推理采用对话续写的形式。
单轮对话
A chat between a human and an artificial intelligence bot. The bot gives helpful, detailed, and polite answers to the human's questions.
|Human|: 你好
|Bot|:
多轮对话
A chat between a human and an artificial intelligence bot. The bot gives helpful, detailed, and polite answers to the human's questions.
|Human|: 你好
|Bot|: 你好,有什么我能帮助你的?<|endoftext|>
|Human|: 介绍下中国的首都吧
|Bot|:
注意:在历史对话的Bot输出结尾,拼接一个<|endoftext|>