You can deploy this model and start earning money today!

RichardErkhov/
stvlynn_-_Qwen-7B-Chat-Cantonese-gguf

Name: stvlynn_-_Qwen-7B-Chat-Cantonese-gguf
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Qwen-7B-Chat-Cantonese - GGUF

Model creator: https://huggingface.co/stvlynn/
Original model: https://huggingface.co/stvlynn/Qwen-7B-Chat-Cantonese/

Name	Quant method	Size
Qwen-7B-Chat-Cantonese.Q2_K.gguf	Q2_K	2.84GB
Qwen-7B-Chat-Cantonese.IQ3_XS.gguf	IQ3_XS	3.23GB
Qwen-7B-Chat-Cantonese.IQ3_S.gguf	IQ3_S	3.32GB
Qwen-7B-Chat-Cantonese.Q3_K_S.gguf	Q3_K_S	3.32GB
Qwen-7B-Chat-Cantonese.IQ3_M.gguf	IQ3_M	3.61GB
Qwen-7B-Chat-Cantonese.Q3_K.gguf	Q3_K	3.78GB
Qwen-7B-Chat-Cantonese.Q3_K_M.gguf	Q3_K_M	3.78GB
Qwen-7B-Chat-Cantonese.Q3_K_L.gguf	Q3_K_L	4.0GB
Qwen-7B-Chat-Cantonese.IQ4_XS.gguf	IQ4_XS	4.02GB
Qwen-7B-Chat-Cantonese.Q4_0.gguf	Q4_0	4.2GB
Qwen-7B-Chat-Cantonese.IQ4_NL.gguf	IQ4_NL	4.22GB
Qwen-7B-Chat-Cantonese.Q4_K_S.gguf	Q4_K_S	4.22GB
Qwen-7B-Chat-Cantonese.Q4_K.gguf	Q4_K	4.56GB
Qwen-7B-Chat-Cantonese.Q4_K_M.gguf	Q4_K_M	4.56GB
Qwen-7B-Chat-Cantonese.Q4_1.gguf	Q4_1	4.62GB
Qwen-7B-Chat-Cantonese.Q5_0.gguf	Q5_0	5.03GB
Qwen-7B-Chat-Cantonese.Q5_K_S.gguf	Q5_K_S	5.03GB
Qwen-7B-Chat-Cantonese.Q5_K.gguf	Q5_K	5.32GB
Qwen-7B-Chat-Cantonese.Q5_K_M.gguf	Q5_K_M	5.32GB
Qwen-7B-Chat-Cantonese.Q5_1.gguf	Q5_1	5.44GB
Qwen-7B-Chat-Cantonese.Q6_K.gguf	Q6_K	5.91GB
Qwen-7B-Chat-Cantonese.Q8_0.gguf	Q8_0	7.65GB

Original model description:

license: agpl-3.0 datasets:

stvlynn/Cantonese-Dialogue language:
zh pipeline_tag: text-generation tags:
Cantonese
廣東話
粤语 base_model: Qwen/Qwen-7B-Chat

Qwen-7B-Chat-Cantonese (通义千问·粤语)

Intro

Qwen-7B-Chat-Cantonese is a fine-tuned version based on Qwen-7B-Chat, trained on a substantial amount of Cantonese language data.

Qwen-7B-Chat-Cantonese係基於Qwen-7B-Chat嘅微調版本，基於大量粵語數據進行訓練。

ModelScope(魔搭社区)

Usage

Requirements

python 3.8 and above
pytorch 1.12 and above, 2.0 and above are recommended
CUDA 11.4 and above are recommended (this is for GPU users, flash-attention users, etc.)

Dependency

To run Qwen-7B-Chat-Cantonese, please make sure you meet the above requirements, and then execute the following pip commands to install the dependent libraries.

pip install transformers==4.32.0 accelerate tiktoken einops scipy transformers_stream_generator==0.0.4 peft deepspeed

In addition, it is recommended to install the flash-attention library (we support flash attention 2 now.) for higher efficiency and lower memory usage.

bash

git clone https://github.com/Dao-AILab/flash-attention
cd flash-attention && pip install .

Quickstart

Pls turn to QwenLM/Qwen - Quickstart

Training Parameters

Parameter	Description	Value
Learning Rate	AdamW optimizer learning rate	7e-5
Weight Decay	Regularization strength	0.8
Gamma	Learning rate decay factor	1.0
Batch Size	Number of samples per batch	1000
Precision	Floating point precision	fp16
Learning Policy	Learning rate adjustment policy	cosine
Warmup Steps	Initial steps without learning rate adjustment	0
Total Steps	Total training steps	1024
Gradient Accumulation Steps	Number of steps to accumulate gradients before updating	8