All stories
11 stories found
Ollama Answers Get Cut Off Mid-Sentence: num_ctx vs num_predict Explained
Truncated replies and models that "forget" the start of a long document come from two different settings. Here is how I tell them apart and the values I use.
Ollama Error "model requires more system memory": How I Fixed It in 10 Minutes
Ollama refuses to load a model and says it needs more system memory than you have. Here is exactly what the number means and the four fixes that worked on my 32 GB / RTX 3060 box.
DeepSeek-R1 Repeats Itself or Outputs Gibberish: The 4 Settings That Fixed It
Endless loops, repeated sentences or mojibake from DeepSeek-R1 are almost always sampling or quantization — not a broken model. The exact parameters I now use.
Ollama Filled My Disk: How I Moved the Models Directory Safely
Ollama stores every model on the system drive by default. The exact steps I used to move ~/.ollama/models to a second drive on Linux, Windows and macOS without re-downloading.
Ollama Pull Fails with "max retries exceeded" or EOF: How I Get Downloads to Finish
Model downloads that stall at 97% or die with EOF. The five things that fixed it for me: resume behaviour, MTU, DNS, proxy variables and a corrupted blob cache.
Ollama "connection refused on 127.0.0.1:11434": The 5 Causes I Have Actually Hit
Your client cannot reach Ollama on port 11434. Here are the five real causes — service not running, wrong host binding, WSL networking, port conflict, and Docker — with the command that fixes each.
DeepSeek-R1 Shows Its <think> Tags in the Output — Here Is How I Strip Them
DeepSeek-R1 prints its whole reasoning chain before the answer. Three ways I remove it: prompt-level, API-level parsing, and a small wrapper I use in every script.
How Much RAM Do You Actually Need for Local AI? I Tested 8, 16, 32 and 64GB
Four machines, same four tasks. 16GB is the real threshold, and the 64GB results were identical to 32GB on everything except one case.
Ollama Running Slow? 7 Fixes That Actually Worked on My Machine
Same laptop, same 7B model: 4 tokens/s before, 31 after. Seven settings were wrong. Here is the exact order I check them in.
Transcribing Audio Locally with Whisper: My Actual Workflow (No Uploads)
Two hours of interview audio transcribed on a laptop in eleven minutes, offline. The model size that is worth it, the flags that matter, and how I handle speaker names.
How to Run DeepSeek-R1 Offline on Mac mini M4 (Step-by-Step 2026 Guide)
Install Ollama, pull the right DeepSeek-R1 distill for your unified memory, and run a private reasoning model on a Mac mini M4 with no internet and no subscription.