the local qwen models are very impressive, even the small 1.7b param qwen3 model doesn't seem stupid like the gemma ones do from my experience.
using opencode rn
listening to olivia rodrigo 🎧
soooo close to 3k followers! https://morel.us-east.host.bsky.network/xrpc/com.atproto.sync.getBlob?did=did:plc:4hawmtgzjx3vclfyphbhfn7v&cid=bafkreibnwbgvwsfscnzmzhn7md5nrt5ofinevx2xq3zodnxmvobau67yjm
just saw this peacock on a random road lmfao https://morel.us-east.host.bsky.network/xrpc/com.atproto.sync.getBlob?did=did:plc:4hawmtgzjx3vclfyphbhfn7v&cid=bafkreihnnmtcxqvvoxbetzertpjlgyucwgby4e65e3f7kuqtbw7imodaqy https://morel.us-east.host.bsky.network/xrpc/com.atproto.sync.getBlob?did=did:plc:4hawmtgzjx3vclfyphbhfn7v&cid=bafkreihylomj36yv2se7ym7dy57rtmud4leik5f43ofut7g7xfs3kkz5ce
i still find it wild that you can download an LLM that runs offline on an iPhone 11 that's more intelligent and can reason better than the current US president 😭
local llms are so fun
c'est la vie
Someone should fork bitchat and make a client that works on old iOS versions. 📱 [github.com/permissionl...]( ) https://morel.us-east.host.bsky.network/xrpc/com.atproto.sync.getBlob?did=did:plc:4hawmtgzjx3vclfyphbhfn7v&cid=bafkreia2gequo5plzoikvayhsa2qz3dcncyzjd7w2ycge5sdwwds4kozfu
Brian May reflects on Queen's performance at Live Aid https://morel.us-east.host.bsky.network/xrpc/com.atproto.sync.getBlob?did=did:plc:4hawmtgzjx3vclfyphbhfn7v&cid=bafkreielllt327oaocu4ug52h4w3qpypik6lxdqold6yb7nvmgitnmcp3u