arcee-ai/Virtuoso-Lite
Text Generation
โข
Updated
โข
117
โข
22
DataSoul
DataSoul
AI & ML interests
AI User
Recent Activity
liked
a model
1 day ago
arcee-ai/Virtuoso-Lite
liked
a model
1 day ago
arcee-ai/Virtuoso-Medium-v2
Organizations
None yet
DataSoul's activity
liked
2
models
1 day ago
reacted to
onekq's
post with ๐
5 days ago
Post
2242
So ๐DeepSeek๐ hits the mainstream media. But it has been a star in our little cult for at least 6 months. Its meteoric success is not overnight, but two years in the making.
To learn their history, just look at their ๐ค repo https://huggingface.co/deepseek-ai
* End of 2023, they launched the first model (pretrained by themselves) following Llama 2 architecture
* June 2024, v2 (MoE architecture) surpassed Gemini 1.5, but behind Mistral
* September, v2.5 surpassed GPT 4o mini
* December, v3 surpassed GPT 4o
* Now R1 surpassed o1
Most importantly, if you think DeepSeek success is singular and unrivaled, that's WRONG. The following models are also near or equal the o1 bar.
* Minimax-01
* Kimi k1.5
* Doubao 1.5 pro
To learn their history, just look at their ๐ค repo https://huggingface.co/deepseek-ai
* End of 2023, they launched the first model (pretrained by themselves) following Llama 2 architecture
* June 2024, v2 (MoE architecture) surpassed Gemini 1.5, but behind Mistral
* September, v2.5 surpassed GPT 4o mini
* December, v3 surpassed GPT 4o
* Now R1 surpassed o1
Most importantly, if you think DeepSeek success is singular and unrivaled, that's WRONG. The following models are also near or equal the o1 bar.
* Minimax-01
* Kimi k1.5
* Doubao 1.5 pro
- 1 reply
Still Failing Merge Attempts
#1 opened 5 days ago
by
DataSoul
updated
a
model
5 days ago
published
a
model
5 days ago
updated
a
model
5 days ago
published
a
model
5 days ago
question
1
#1 opened 5 days ago
by
DataSoul
reacted to
lewtun's
post with ๐ฅ
5 days ago
Post
9492
We are reproducing the full DeepSeek R1 data and training pipeline so everybody can use their recipe. Instead of doing it in secret we can do it together in the open!
๐งช Step 1: replicate the R1-Distill models by distilling a high-quality reasoning corpus from DeepSeek-R1.
๐ง Step 2: replicate the pure RL pipeline that DeepSeek used to create R1-Zero. This will involve curating new, large-scale datasets for math, reasoning, and code.
๐ฅ Step 3: show we can go from base model -> SFT -> RL via multi-stage training.
Follow along: https://github.com/huggingface/open-r1
๐งช Step 1: replicate the R1-Distill models by distilling a high-quality reasoning corpus from DeepSeek-R1.
๐ง Step 2: replicate the pure RL pipeline that DeepSeek used to create R1-Zero. This will involve curating new, large-scale datasets for math, reasoning, and code.
๐ฅ Step 3: show we can go from base model -> SFT -> RL via multi-stage training.
Follow along: https://github.com/huggingface/open-r1
- 5 replies
updated
a
model
8 days ago
published
a
model
8 days ago
updated
a
model
8 days ago
published
a
model
8 days ago
liked
a
model
9 days ago
Produces Garbige
6
#3 opened 9 days ago
by
Nycoorias
liked
a
model
9 days ago
SFT (Non-RL) distillation is this good on a sub-100B model?
3
#2 opened 11 days ago
by
KrishnaKaasyap
updated
a
model
11 days ago
liked
a
model
11 days ago