Internet man with personal thoughts. Pseudo scientist through Gฬถoฬถoฬถgฬถlฬถeฬถ AI.
30.83001,108.355886
Joined May 2008
- Tweets30.5K
- Following101
- Followers1.4K
- Likes943
ๅชๆณ่ฏด๏ผ AI ็ๆฌ @deepseek_ai ไธไธช่ฎข้
๏ผ
ๅทฅไธๆถไปฃๅผ้ขไบบ็ฑป่ฟๅ็ปฟ่ฒ่ฝๆบ็่ฟๆฏ @Tesla ๏ผๆฒกๆณๅฐไฟกๆฏๆถไปฃ็ฌฌไธไธชๅๅบ่ฏฅๅฃๅท็ๆฏ @deepseek_ai ใไธ็ฅ้ @elonmusk ไฝไฝๆๆณ๏ผ
ๅทฅไธๆถไปฃๅผ้ขไบบ็ฑป่ฟๅ็ปฟ่ฒ่ฝๆบ็่ฟๆฏ @Tesla ๏ผๆฒกๆณๅฐไฟกๆฏๆถไปฃ็ฌฌไธไธชๅๅบ่ฏฅๅฃๅท็ๆฏ @deepseek_ai ใไธ็ฅ้ @elonmusk ไฝไฝๆๆณ๏ผ
๐ Introducing DeepInfer, our in-house inference framework, now fully open source!
Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. Weโre sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable.
๐ github.com/deepseek-ai/DeepIโฆ
1/4
๐๐๐๐๐๐๐๐ retweeted
DeepSeek just open sourced its high-performance, large-scale inference framework DeepInfer in Apache-2.0 license.
github.com/deepseek-ai/DeepIโฆ
Supported features:
- Optimized kernels, model and sampler runtime
- Parallelism: Attention DP / balanced CP, MoE EP + EPLB
- DSpark and confidence-scheduled verification
- SWA bounded replay
- Asynchronous scheduling (zero-overhead batch scheduler)
- Batch-invariant inference
- Deterministic sampling
- Structured output
I just cooked up a new ChatGPT MCP app all by itselfโand it CAN make GPT great again! Try it at studiocat.app.
#X ๆไนๆ่ฟๆปๆฏ็ปๆๆจ้ธกๆฑคๆ๏ผไธ็ฅ้ๆ่ฏปไนฆๆถไบบ้ๅคๅทโๆค้โๅ๏ผๅข๏ผ้ฃๆฏๅฎ่ฟไธๅญๅจ๐ค
#ChatGPT is still just too powerfulโit actually supports Subagents on the web too!
#Codex ๅณไฝฟ instructions ่ฆๆฑๅ
ถ็จไธญๆๅๅค๏ผๅฎไนไผๅจ่ฟญไปฃๅ ่ฝฎๆ subagents ๆปๅๆขๅ่ฑๆ๏ผๆไปฅไป็ๆไปค้ตไปๆฏๆไนๆตๅพ้ซๅ็๏ผ
ๆฒกๆณๅฐๆ่ฟ้ๅนด่ๆง็ iPhone XR ็ซ็ถ่ฟ่ฝ็จ๏ผ่นๆ็็ๆฏๅถ้ ไธ็่ฏๅฟๅๅฎถ๏ผ
ๆฏไปฅไธบ #ChatGPT ็ MCP Tunnel ่ฎพ่ฎก็็ๆฏ้ขๅ
ๅๅ๏ผไป
้่ฟๅบ็ซ้้ๅฎ็ฐ็จๆทๅฐๆฌๅฐๆฅๅ
ฅ AI Exec๏ผๅ
ผ้กพไพฟๆทๆงไธๅฎๅ
จๆง๏ผๆไปฅ่ฏดๅจ่ง่ไธ A\ ่ตฐๅจๅ้ข๏ผๅจๅทฅ็จๆๆฏไธ OpenAI ้ฅ้ฅ้ขๅ
ใ
#Codex ็ Subagent ๆง่ก็ปๅธธไผ้ๅฐๅฎๅ
จๅฎกๆฅ็ๆฆๆช๏ผ็ถๅๆๆญๆดไธชๆง่ก๏ผ
cc @thsottiaux
