葉佐俊
win10
·
AI & ML interests
None yet
Organizations
win10's activity
使用多少token訓練
#1 opened 2 days ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Will it be available in multiple languages?
1
#1 opened 15 days ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
這是我合併的13B基礎模型(7B到13b)
2
#2 opened 16 days ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
想請問一下會有int4版本嗎?
#2 opened about 1 month ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Can you add an fp8 or int4 quantization loader?
1
#3 opened about 1 month ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
What do you think is better out of the box?
#1 opened about 1 month ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
taide-meta-it-16b合併模型
#10 opened about 1 month ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Great Model
15
#1 opened about 2 months ago
by
spike4379
mixtral format?
5
#1 opened about 1 month ago
by
KnutJaegersberg
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1669551186189-63732ebbbd81fae2b3aaf3fb.jpeg)
提議:官方放到ollama上
2
#1 opened about 2 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Hi! Friends in the open source community, will you open source?
4
#1 opened 4 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Hi, will there be an instruction version of the coding model in the future?
#2 opened 4 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
这是很棒的模型但有个问题
1
#2 opened 4 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
之後會有從頭開始預訓練的模型嗎?
2
#1 opened 5 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Can you please fine-tune the RWKV-5 7B?
1
#2 opened 5 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Will there be a version with traditional Chinese in the future?
#5 opened 5 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Can I ask you to try training TinyDolphin-2.8-1.1b or create multilingual data?
1
#9 opened 5 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
A more complex script generated by gpt4-turbo
1
#8 opened 5 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
The number of samples is 5,000,000, from mathrandom2.py.
1
#6 opened 5 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
A more complex script generated by gpt4-turbo
1
#3 opened 5 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
A more complex script generated by gpt4-turbo
#4 opened 5 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
A more complex script generated by gpt4-turbo
#2 opened 5 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
您是怎么合并模型的呢?
#2 opened 6 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Can you fine tune tinyllama1.1b-3t?
3
#1 opened 6 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
You should try training a model with 2B parameters and context length 32000.
1
#3 opened 6 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
您们能尝试训练开源的芯片生成模型吗?
#2 opened 6 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
会有新的模型吗?
6
#1 opened 6 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Will there be a Chinese version of the model later?
#2 opened 6 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
How did you merge the models?
5
#1 opened 6 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Please, you make a 6x1b moe model?
1
#1 opened 6 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Can you make a model that can run without quantization on a gpu with only 8g vram?
1
#2 opened 6 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
What data set did you use?
#1 opened 6 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Nice work!
7
#3 opened 7 months ago
by
BrainSlugs83
Request for Open Sourcing a Chat-Chip Model for Chip Design
#10 opened 7 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
hello, I would like to ask you how to merge the 20B model?
12
#1 opened 8 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
How to convert to gguf format?
#5 opened 8 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Can you do the same thing with Yi-6b-200k?
3
#6 opened 8 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
How to convert the llama2.c model into huggingface model format?
2
#4 opened 9 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
您能嘗試使用开放源码訓練模型嗎?
2
#2 opened 9 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
您好能打包成zip文件嗎?
1
#2 opened 10 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
有增量訓練分子生成或創造模型的計畫
5
#1 opened 10 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
New activity in
xiaol/RWKV-toolformer-translation-japanese-chinese-english-7B-World-128k
10 months ago
您們能嘗試增強RWKV的編程能力嗎?
1
#1 opened 10 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
你好,你能嘗試在保留模型原本能力的同時,訓練模型能生成化學分子與分析嗎?
7
#1 opened 11 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Can this dataset support effective 65K contextual training?
#2 opened 11 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Can try code as long text data.
1
#1 opened 11 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Your 3b model is very exciting and proves that data improvement works!
#13 opened 11 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Hi, can you do the same for the rwkv-world model?
#4 opened 11 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
[Digression] Hello, do you plan to fine-tune the rwkv-world model?
1
#3 opened 11 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Will the team consider using RWKV-world to train new models in the future?
#1 opened 12 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Is there a merge model script?
#2 opened 12 months ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Hello, how to do lora fine-tuning in oobabooga/text-generation-webui.
1
#1 opened about 1 year ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Using an embedding space, solve the context length problem?
#16 opened about 1 year ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Hello, can the rtx 3050 8g run this model?
1
#1 opened about 1 year ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Consider using the MPT-7b model to train the new pygmalion model?
1
#9 opened about 1 year ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
能用.safetensors格式加载chatglm-6b-int4吗?
#9 opened about 1 year ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Adding `safetensors` variant of this model
#6 opened about 1 year ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)
Add ggml-model-bloom-3b-zh-f16-q4_0.bin from bloomz.cpp converter.
1
#2 opened about 1 year ago
by
win10
![](https://cdn-avatars.huggingface.co/v1/production/uploads/1678188568629-noauth.png)