Lighting AI is so underrated. I’ve been using it for months now since you first mentioned it. It’s freaking awesome! I try to use it as my power PC when I can’t do things on my personal 8GB GPU 🥲
Love all your free videos and all the options you provide. Would love to see some additional app building videos. Also interested in long time low-cost deployment options, so are dokploy and coolify still the way to go etc :)
It seems silly but the only reason that I am still using Cursor is that its Autocompletion feature (TAB) still seems better than other alternatives: github copilot, continue, supermaven and others. It's not relevant to the LLM models but to the way how it works. Hope these free vscode plugins may also improve on this. ^_^
you know what would be reallll clutch and cool!? if you had a simple video explaining how to make a "base setup of tools" that you can use when moving from like codespaces to other places like idx or lightningai you can bring your own config im pretty sure like maybe a dev container with aider a cool starship promt .zsh i reckon with somekind of advanced syntax highlighting and autocomplete (having a hard rime finding something that works everywhere i want it in terminal and nvim/cursor/vscode does supermaven work in terminal without nvim?)
Are there downsides for Lightning? Going through it, it seems way better for ML projects than IDX and CodeSpace, as far as Docker integration, custom apt-get system installs, code sharing, and GPU customization. I'm not sure why anyone wouldn't want to use it. Should I wait for the other shoe to drop?
It's really great and the VS Code setup of theirs seems to like real VS Code instead of a crappy version that Codespace has (despite being from microsoft themselves). The only issue might be that it's limited to One studio but it's fine for me. It also has 16gigs of Ram which I didn't say in the video but that's pretty significant as well. I haven't found a downside yet. But, I have seen some users complaining about their pricing of GPUs. I don't know how true is that.
@@AICodeKing For clients/devs looking to collaborate and/or run parts of projects, this is way better than trying to get them to install a local Docker. At the early outset of using it, I think $50/month is a good deal, especially if you can scale the RAM and GPU (pay as you go) whenever you want.
Can you make a video on how to use supermaven alone. I added a credit card and now have 1 free month trial of super maven but it doesn't seem to work like cline? Does this mean I have to create the files and copy and paste the code?
thanks again , please i love how idx google works with the gemini in the config u can make it detect ur codebase context but i dont like how google idx when i start new project it just remove all my extensions profiles and everythin and start again from zero , i tried to export default profile and its not work , can u make video for how to use google idx , thanks i am oldest and top fan of ur channel
@@AICodeKing I can't use any model except qwen2.5 tool 7b and 14b. All of them give this error except for it and I don't know how to overcome it. maybe it's because I don't speak English. I don't even know
@@singsingleung8187 no, I don’t pretend to be Claude, I just use ollama and only qwin works, the rest don’t. literally this is an error. I can't solve it. Cline uses complex prompts and iterative task execution that may be challenging for less capable models. For best results, it's recommended to use Claude 3.5 Sonnet for its advanced agentic coding capabilities.
@@singsingleung8187 no, I don’t pretend to be Claude, I just use ollama and only qwin works, the rest don’t. literally this is an error. I can't solve it. Cline uses complex prompts and iterative task execution that may be challenging for less capable models. For best results, it's recommended to use Claude 3.5 Sonnet for its advanced agentic coding capabilities.
ahahahha somebody ask, why you create basics apps, guys its a 12 min video, what do you want come on? do it by yourself, the importance is the tools and how to connect all
bro is on a grind, saw his videos and bro posting videos like he having lunch!
Lighting AI is so underrated. I’ve been using it for months now since you first mentioned it. It’s freaking awesome! I try to use it as my power PC when I can’t do things on my personal 8GB GPU 🥲
So you subscribed for the service right?
u use lightning ai's gpu for what?
Love all your free videos and all the options you provide. Would love to see some additional app building videos. Also interested in long time low-cost deployment options, so are dokploy and coolify still the way to go etc :)
It seems silly but the only reason that I am still using Cursor is that its Autocompletion feature (TAB) still seems better than other alternatives: github copilot, continue, supermaven and others. It's not relevant to the LLM models but to the way how it works. Hope these free vscode plugins may also improve on this. ^_^
Create an analysis app, like RAG, but need AI to analyze the result
you know what would be reallll clutch and cool!? if you had a simple video explaining how to make a "base setup of tools" that you can use when moving from like codespaces to other places like idx or lightningai you can bring your own config im pretty sure like maybe a dev container with aider a cool starship promt .zsh i reckon with somekind of advanced syntax highlighting and autocomplete (having a hard rime finding something that works everywhere i want it in terminal and nvim/cursor/vscode does supermaven work in terminal without nvim?)
goddamn it brother, put a damn comma in there.
@@sleeplessfeed 💀 good ole fashion schizo rant lol
Are there downsides for Lightning? Going through it, it seems way better for ML projects than IDX and CodeSpace, as far as Docker integration, custom apt-get system installs, code sharing, and GPU customization. I'm not sure why anyone wouldn't want to use it. Should I wait for the other shoe to drop?
It's really great and the VS Code setup of theirs seems to like real VS Code instead of a crappy version that Codespace has (despite being from microsoft themselves). The only issue might be that it's limited to One studio but it's fine for me. It also has 16gigs of Ram which I didn't say in the video but that's pretty significant as well. I haven't found a downside yet. But, I have seen some users complaining about their pricing of GPUs. I don't know how true is that.
@@AICodeKing For clients/devs looking to collaborate and/or run parts of projects, this is way better than trying to get them to install a local Docker. At the early outset of using it, I think $50/month is a good deal, especially if you can scale the RAM and GPU (pay as you go) whenever you want.
yes it's pretty cool, i would like something easy to set up and cheaper than bolt
stackblitz with cline.
Can you make a video on how to use supermaven alone. I added a credit card and now have 1 free month trial of super maven but it doesn't seem to work like cline? Does this mean I have to create the files and copy and paste the code?
thanks again , please i love how idx google works with the gemini in the config u can make it detect ur codebase context but i dont like how google idx when i start new project it just remove all my extensions profiles and everythin and start again from zero , i tried to export default profile and its not work , can u make video for how to use google idx , thanks i am oldest and top fan of ur channel
Great!
Bro , test Pc agent by Mobile agent
It is awesome for controlling computer.
Plz plz do it soon 🙏🙏🙏
How to install flux 1 dev in lightning AI?
At the end of your video ollama gives you an error, how can I deal with it?
It's an error because I just said "Hi"... And, Cline wants to do change files and stuff but I rejected it. It's not an Ollama error.
@@AICodeKing I can't use any model except qwen2.5 tool 7b and 14b. All of them give this error except for it and I don't know how to overcome it. maybe it's because I don't speak English. I don't even know
@alex_great23 Same issue, seem claude dev promblem
@@singsingleung8187 no, I don’t pretend to be Claude, I just use ollama and only qwin works, the rest don’t. literally this is an error. I can't solve it. Cline uses complex prompts and iterative task execution that may be challenging for less capable models. For best results, it's recommended to use Claude 3.5 Sonnet for its advanced agentic coding capabilities.
@@singsingleung8187 no, I don’t pretend to be Claude, I just use ollama and only qwin works, the rest don’t. literally this is an error. I can't solve it. Cline uses complex prompts and iterative task execution that may be challenging for less capable models. For best results, it's recommended to use Claude 3.5 Sonnet for its advanced agentic coding capabilities.
First comment bro❤❤
ahahahha somebody ask, why you create basics apps, guys its a 12 min video, what do you want come on? do it by yourself, the importance is the tools and how to connect all