hiro
hiro
# Problem - Currently Jan automatically get the `logical`/ `physical` CPU count (in case the machine does not have GPU) for model inference, which causes inefficiency. - If it's too...
**Problem** - In long conversation especially with file uploaded, it's likely that the LLM cannot remember the conversational history which is long away (underlying this exceeds LLM context window). **Success...
**Problem** - While chatting with my local file, I find it useful if I can see the references that LLM uses in order to answer me - See the example...
**Problem Link:** Check out the issue on GitHub: [Issue #2432](https://github.com/janhq/jan/issues/2432). **Why This Matters:** Jan is designed to work best with newer technology, using something called AVX2 for processing. However, if...
**Describe the bug** - Context: https://discord.com/channels/1107178041848909847/1222084065918713909 - User though do not have NVIDIA CUDA (they only have small AMD GPU - `Rx580`) but they are forced to use `win-cuda-12-0` which...
**Describe the bug** - If `Retrieval` is ON, even though I upload `image` it does not understand the image content. - Upon checking the network, the image content is dropped...
**Problem** - Context: https://discord.com/channels/1107178041848909847/1222681742712897609 - User with old mac want to use Jan with AMD Radeon GPU, however the current vulkan does not support it **Success Criteria** - Just work,...
**Problem** - Context: https://discord.com/channels/1107178041848909847/1223957838678724650/1223957838678724650 - Jan app currently does not support nicely for JP/ CN language while LLM is all about language - This has negative effect to Jan goal...
- [x] Update `nitro-set-version` with cortex - [x] Remove metal file in mac-arm64
**Problem** - Context: https://discord.com/channels/1107178041848909847/1222681742712897609 - User with Mac Intel and AMD Radeon wants to use Vulkan to accelerate models **Success Criteria** - Nitro build to support this **Additional context** Add...